Commodity
Derivatives
Founded in 1807, John Wiley & Sons is the oldest independent publishing company in
the United States. With offices in North America, Europe, Australia and Asia, Wiley
is globally committed to developing and marketing print and electronic products and
services for our customers’ professional and personal knowledge and understanding.
The Wiley Finance series contains books written specifically for finance and investment professionals as well as sophisticated individual investors and their financial
advisors. Book topics range from portfolio management to e-commerce, risk management, financial engineering, valuation and financial instrument analysis, as well as
much more.
For a list of available titles, visit our Web site at www.WileyFinance.com.
Commodity
Derivatives
Markets and Applications
NEIL C. SCHOFIELD
Second Edition
This edition first published 2021
Copyright © 2021 by Neil C. Schofield.
Registered office
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ,
United Kingdom
For details of our global editorial offices, for customer services and for information about how to apply
for permission to reuse the copyright material in this book please see our website at www.wiley.com.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of
the publisher.
Wiley publishes in a variety of print and electronic formats and by print-on-demand. Some material
included with standard print versions of this book may not be included in e-books or in print-on-demand.
If this book refers to media such as a CD or DVD that is not included in the version you purchased, you may
download this material at http://booksupport.wiley.com. For more information about Wiley products, visit
www.wiley.com.
Designations used by companies to distinguish their products are often claimed as trademarks. All brand
names and product names used in this book are trade names, service marks, trademarks or registered
trademarks of their respective owners. The publisher is not associated with any product or vendor
mentioned in this book.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts
in preparing this book, they make no representations or warranties with respect to the accuracy or
completeness of the contents of this book and specifically disclaim any implied warranties of
merchantability or fitness for a particular purpose. It is sold on the understanding that the publisher is
not engaged in rendering professional services and neither the publisher nor the author shall be liable
for damages arising herefrom. If professional advice or other expert assistance is required, the services
of a competent professional should be sought.
Library of Congress Cataloging-in-Publication Data
Names: Schofield, Neil C., author. | John Wiley & Sons, publisher.
Title: Commodity derivatives : markets and applications / Neil C.
Schofield.
Description: Second edition. | [Hoboken, NJ] : Wiley, 2021. | Includes
index.
Identifiers: LCCN 2021002135 (print) | LCCN 2021002136 (ebook) | ISBN
9781119349105 (hardback) | ISBN 9781119349228 (adobe pdf) | ISBN
9781119349259 (epub)
Subjects: LCSH: Commodity futures. | Commodity exchanges. | Derivative
securities.
Classification: LCC HG6024.A3 S364 2021 (print) | LCC HG6024.A3 (ebook) |
DDC 332.64/57—dc23
LC record available at https://lccn.loc.gov/2021002135
LC ebook record available at https://lccn.loc.gov/2021002136
Cover Design: Wiley
Cover Images: Oil refinery © Thatree Thitivongvaroon/
Moment Open/Getty Images, Airplane © cate_89/Shutterstock,
Copper rods © Aksenenko Olga/Shutterstock,
Silver © BestPix/Shutterstock,
Gold bars © boykung/Shutterstock
Set in 10/12pt STIXTwoText by SPi Global, Chennai, India
Printed in
10 9 8 7 6 5 4 3 2 1
TO REGGIE, BRENNIE, ROBERT, AND GILLIAN
TO NICKI
Contents
Preface
xi
CHAPTER 1
Fundamentals of Commodities and Derivatives
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
Market overview
Market participants
Traded versus non-traded commodities
Forward contracts
Futures
Swaps
Options
Exotic options
CHAPTER 2
Derivative Valuation
2.1
2.2
2.3
2.4
2.5
2.6
2.7
Asset characteristics
Commodity prices and the economic cycle
Principles of commodity valuation
Forward price curves
Commodity swap valuation
Principles of option valuation
Measures of option risk management
CHAPTER 3
Risk Management Principles
3.1
3.2
3.3
3.4
3.5
3.6
Defining risk
Commodity market participants – the time dimension
Hedging corporate risk exposures
A framework for analysing corporate risk
Hedging customer exposures
Trading risk management
CHAPTER 4
Gold
4.1
4.2
1
2
3
6
8
8
10
11
14
18
18
18
19
21
32
39
44
58
58
60
61
62
63
66
76
The market for gold
Gold price drivers
76
81
vii
viii
CONTENTS
4.3
4.4
4.5
4.6
4.7
The gold leasing and deposit market
Hedging
Trading gold
Yield enhancement
Summary
91
96
106
111
112
CHAPTER 5
Base Metals
114
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
5.10
114
115
119
120
122
130
133
134
139
157
Overview of base metal production
The copper lifecycle
Aluminium
The Steel market
The London Metal Exchange
Base metal price drivers
Electric Vehicles
Structure of market prices
Hedges for aluminium consumers in the automotive sector
Summary
CHAPTER 6
Crude Oil
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
159
Overview of energy markets
The value of crude oil
An overview of the physical supply chain
Refining crude oil
The demand for and supply of crude oil
Price drivers
The price of crude oil
Trading crude oil and refined products
Notes
Managing price risk along the supply chain
159
159
163
165
172
179
189
195
209
212
CHAPTER 7
Natural Gas
237
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
237
238
238
242
245
250
257
260
263
Formation of natural gas
Measuring natural gas
The physical supply chain
Deregulation and re-regulation
The demand for and supply of natural gas
Natural gas prices
Natural gas price drivers
Trading natural gas
Natural gas derivatives
ix
Contents
CHAPTER 8
Electricity
280
8.1
8.2
8.3
8.4
8.5
8.6
280
283
285
291
301
313
What is electricity?
The physical supply chain
Market structure and regulation
Price drivers of electricity
Trading electricity – an overview
Electricity derivatives
CHAPTER 9
Plastics
9.1
9.2
9.3
9.4
9.5
9.6
9.7
9.8
9.9
9.10
332
The chemistry of plastic
The production of plastic
Monomer production
Polymerisation
Applications of plastics
Summary of the plastics supply chain
Price determination
Plastic price drivers
Forwards and swaps
Option strategies
CHAPTER 10
Bulk Commodities
10.1
10.2
10.3
10.4
10.5
10.6
The basics of coal
The demand for and supply of coal
Coal – the physical supply chain
Coal derivatives
Iron ore
Freight markets – the fundamentals
CHAPTER 11
Climate and Weather
11.1
11.2
11.3
11.4
11.5
11.6
11.7
11.8
The science of climate change
The consequences of climate change
The argument against climate change
History of human action against climate change
Price drivers of emissions markets
EU Emission Trading System
Emission derivatives
Weather derivatives
332
333
334
334
335
336
336
337
338
340
341
341
343
346
349
353
355
368
368
371
372
373
376
379
384
388
x
CONTENTS
CHAPTER 12
Agriculture
394
12.1
12.2
12.3
12.4
12.5
12.6
12.7
12.8
394
394
395
403
409
412
421
426
Agricultural markets
Definitions
Agricultural products
Soft commodities
Ethanol
Price drivers
Exchange traded agricultural and ethanol derivatives
Over-the-counter agricultural derivatives
CHAPTER 13
Commodity-Linked Financing
13.1
13.2
13.3
13.4
The financing need
Project finance
Working capital and the asset conversion cycle
Longer-term debt funding solutions
CHAPTER 14
Commodity Investing
14.1
14.2
14.3
14.4
14.5
14.6
14.7
14.8
Commodity investors
Preferred instruments
Market size
Rationale for investing in commodities
Commodity indices
Total Return Swaps
Exchange traded products (ETPs)
Structured products
433
433
437
440
454
463
463
465
465
466
469
478
481
487
Glossary
499
Bibliography
507
Biography
510
Index
511
Preface
S
ince the start of this century, the commodity markets have fallen in and out of favour
with the financial community. Throughout the preparation of the manuscript for the
second edition, I would often see headlines that made me wonder whether my target
audience would still exist by the time of publication! Having been in finance for over
30 years, one thing I have seen is the way in which history tends to repeat itself, so
fingers crossed.
My original motivation for writing the book, however, stemmed from my time
working at Barclays Investment Bank, where I had tremendous difficulty in finding
people who could provide classroom training on the various commodity products.
Although many companies were able to provide training that described the physical
market for each commodity, virtually no one provided training on over-the-counter
(OTC) derivative structures. As they say, if you want a job done properly . . .
While doing research for the first edition I felt that much of the available documentation either had a very narrow focus concentrating on just one product, or were general
texts on trading commodity futures with a lot of coverage of subjects like technical analysis with little insight into the underlying markets. As a result, I have tried to write a book
that documents in one place the main commodity markets and their associated derivatives.
Within each chapter, I have tried to keep the structure fairly uniform. Typically,
there will be a short section explaining what the commodity is in non-technical terms.
For those with a background in any one specific commodity, this may appear somewhat simplistic, but is included to ensure a reader has sufficient background to place
the subsequent discussion within some context. Typical patterns of demand and supply
are considered as well as the main factors that will influence the price of the commodity.
The latter part of each chapter focuses on the physical market of the particular commodity before detailing the main exchange traded and OTC products.
One of the issues I faced when writing each chapter was to determine which products should be included. I was concerned that I might end up repeating ideas that had
been covered in earlier chapters. Therefore, I have tried to document structures that
are unique to each market in each particular chapter, while the more generic structures
have been spread throughout the text.
So why a second edition? While considering the prospect of writing a second edition, I came across this wonderful quote.
‘Writing as a profession is a sequence of failed ambitions. You never succeed
in writing the book you want, or you’d never bother to write the next. So for
writers, ambition is irrelevant. All you can do is write as well as your talent
will allow’.
—Faye Weldon, Financial Times, weekend magazine, 28/29 July 2012
xi
xii
PREFACE
Upon reflection, I pondered whether the first edition achieved a good balance
between the markets and the derivatives, and so I decided to include more example
transactions. I also took the opportunity to update the original text and expand
the product coverage. For those of you who are wondering if it is worth upgrading
to the second edition, here is a quick snapshot of the major changes. The first edition
ran to about 100,000 words, while this shiny new version is twice the size.
New topics include:
▪ Dual curve swap valuation.
▪ Option valuation within a negative price environment (the Bachelier model).
▪ Volatility skews, smiles, smirks, and term structures for the major commodities.
▪ Case studies on corporate failures linked to commodity risk management.
▪ Implications of growing interest in electric vehicles on commodity markets.
▪ Increased coverage of oil refiners and the challenge of output optimisation.
▪ Expanded sections on the Brent and WTI physical markets.
▪ Expanded section on the trading of electricity.
▪ Inclusion of iron ore and freight markets alongside coal in a chapter renamed as
‘bulk commodities’.
▪ New section on weather derivatives.
▪ Expanded content in the agriculture chapter.
▪ New chapter on commodity financing covering areas such as project finance, working capital management, and commodity-linked debt structures.
▪ Significant rewriting of the chapter on investment structures with new content
illustrating why the concept of ‘roll yield’ is widely misunderstood.
Chapter 1 provides an overview of commodity markets and then outlines the
main derivative building blocks. Chapter 2 considers the main principles of derivative
valuation. This sets the scene for a discussion on the concept of risk management
in Chapter 3. Two different perspectives are taken, that of a corporate with a desire
to hedge some form of exposure and an investment bank that will take on the risk
associated by offering any solution. Chapter 4 looks at the market for gold while
Chapter 5 develops the metal theme to cover base metals. Some readers may complain
that there is no coverage of other precious metals such as silver, platinum, and
palladium, but I felt that including sections on these metals would amount to overkill
and that gold was sufficiently interesting in itself to warrant an extended discussion.
The next three chapters cover the core energy markets, the first of which is crude
oil in Chapter 6. Chapter 7 covers natural gas markets while Chapter 8 looks at
electricity. Chapter 9 describes the market for plastic, which at the time of the first
edition was ‘the new kid on the block’. This has been rewritten to reflect how it has
changed in the subsequent years. Chapter 10 has been renamed as ‘bulk commodities’
with the coverage widening to include iron ore and freight. Chapter 11 looks at the
continuing interest in the trading of carbon emissions but also includes a discussion
on weather derivatives. Chapter 12 covers agricultural products and has been
expanded to cover more markets and a greater number of transactions. Chapter 13
is new and looks at different aspects of commodity finance. The book concludes by
looking at the use of commodities within an investment portfolio.
Preface
xiii
As ever, it would be arrogant of me to assume that this was entirely my own work.
The book is dedicated to the late Paul Roth, who was taken from us far too early in life. In
the decade that I knew him, I benefited considerably from his insight into the world of
derivatives. It never ceased to amaze how after days of pondering on a problem, I could
half explain something I half understood to him and he would be able to explain it back
to me perfectly in simple and clear terms.
Thanks to the team at Wiley who were very patient and understanding while I was
preparing the manuscript. I am sorry it took so long.
General thanks go to my late father Reg Schofield who offered to edit large chunks
of the original manuscript to tidy up ‘the English what I wrote’. Rachel Gillingham
deserves a special mention for helping me express the underlying chemistry of a number of commodities within the book. Her input added considerable value to the overall
manuscript.
At Barclays Capital I would like to thank Arfan Aziz, Natasha Cornish, Lutfey
Siddiqui, Benoit de Vitry, and Troy Bowler. They all endured endless requests for
help and have given generously of their time without complaint. In relation to
specific chapters, thanks go to Matt Schwab and the late Jon Spall (Gold); Angus
Mcheath, Frank Ford, and Ingrid Sternby (Base Metals); David Paul and Nick Smith
(Plastics); Thomas Wiktorowski-Schweitz, Orrin Middleton, Suzanne Taylor, and
Jonathon Taylor (Crude Oil); Simon Hastings, Rob Bailey, David Gillbe (Electricity);
Paul Dawson and Rishil Patel (Emissions); Rachel Frear and Marco Sarcino (Coal);
Maria Igweh (Agricultural). Thanks also to Steve Hochfeld who made some valuable
comments on the agricultural chapter.
With respect to the second edition, John Fry and Neil Scurlock at ICE were generous in their time helping me out with crude oil EFPs. All of them contributed fantastic
insights into the different markets and often reviewed drafts of the manuscript, which
enhanced it no end.
Very special thanks to Nicki, who never once complained about the project and has
always been very interested and supportive of all I do.
If I have missed anyone, then please accept my apologies, but rest assured I am
grateful. Although I did receive a lot of help in compiling the materials, any mistakes
that are in the text are entirely my responsibility.
I am always interested in any comments or suggestions on the text and I can be
contacted at:
neil.schofield@fmarketstraining.com
Neil C. Schofield
P.S. Hi to Alan and Roger, who dared me to include their names. You still owe me
tea and toast from the first edition!
CHAPTER
1
Fundamentals of Commodities
and Derivatives
A
fter the publication of the first edition of this text, many of the author’s friends not
involved with financial markets often asked, ‘what are commodities’? Like many
innocent questions, they are often very difficult to answer. In one sense, they are largely
unprocessed or semi-processed goods, which are either consumed or can be processed
and then resold. However, this definition will not always universally apply; for example,
freight and carbon emission markets do not easily fall within this category.
In general terms, commodities can be classified under different headings:
Energy markets
▪ Crude oil and refined products (e.g. WTI/Brent, gasoline)
▪ Power and natural gas
▪ Natural gas liquids (e.g. propane and butane)
▪ Coal
Industrial metals
▪ Copper, aluminium
Precious metals
▪ Gold, silver
Agricultural products
▪ Grains
▪ Softs (e.g. coffee)
▪ Livestock
‘Specialty’ markets
▪ Forest products (e.g. pulp and recovered paper)
▪ Carbon emissions
▪ Weather
▪ Freight
1
2
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
1.1
MARKET OVERVIEW
Figure 1.1 is a ‘big picture’ overview of commodity markets.
In this diagram there are two main segments, the physical and the financial markets. The diagram was designed without a specific product in mind, but if the reader
prefers some context, it may be helpful to think of a popular commodity such as crude
oil. Within the physical side of the market there will be three main participants: producers, refiners, and consumers. In addition, trading houses will perform a variety of
tasks, which are detailed in a subsequent section. The financial side of the market will
incorporate those entities offering financing and risk management services as well as
investors seeking to earn a return from the asset class. One aspect that is central to
commodities is price discovery, and so the role of futures exchanges is key.
To get a sense of the generic market flows associated outlined in Figure 1.1, consider
the following issues faced by market participants:
▪ Commodities are not homogeneous – it is not particularly helpful to speak in general
terms about commodities. For example, the phrase ‘crude oil’ is meaningless as the
chemical properties of crude extracted in one location will vary from those in a
different location. Trafigura (2016) argues that over 150 types of crude oil are traded
worldwide.
TRADING HOUSES
Risk management
HEDGE FUNDS
INSTITUTIONAL INVESTORS
Strategic commodity
exposure
View driven strategies
FINANCIAL MARKET
COMMERCIAL BANKS
• Own physical assets
• Financing
• Risk management
FUTURES
EXCHANGE
• Benchmark price
• Risk management
• Source of supply
REFINERS
PRODUCERS
Commercial contracts
PHYSICAL MARKET
FIGURE 1.1 Commodity market overview.
CONSUMERS
Commercial contracts
TRADING HOUSES
• Own physical assets
• Trade physical commodity
• Risk management
Fundamentals of Commodities and Derivatives
3
▪ Commodities need to be transformed into consumer goods – for example, oil needs to
be refined to produce gasoline.
▪ Benchmarks help participants agree on a price for non-homogeneous products – so
with respect to crude oil, a particular grade of oil could be priced relative to an
agreed benchmark such as a futures contract that references Brent Blend.
▪ Production and consumption may not take place in the same geographical location – this means that there is a need for transportation. The mode of this
transportation can vary for a single commodity. For example, in the USA, crude oil
is typically moved by pipeline or train. In other areas such as Europe, sea-borne
transport may be more common.
▪ Consumption and production may not occur simultaneously – a consumer may not
need to take immediate delivery of a commodity, therefore storage and inventories
are key factors. When there is a geographic element to the issue, it takes time for a
commodity to be transported.
1.2
MARKET PARTICIPANTS
Market participants are able to manage the respective price risks using derivatives.
Although risk management will be considered in greater detail in Chapter 3, it is worth
considering some related motivations.
Participants can:
▪ Avoid risk,
▪ Retain risk,
▪ Transfer risk,
▪ Reduce risk,
▪ Increase risk.
One of the key roles of derivatives is that they allow different market participants
with different risk profiles and objectives to obtain a desired risk exposure. With respect
to commodity derivatives the main participants will be physical market participants,
price reporting agencies (PRAs), investment banks, commodity trading houses, hedge
funds, or ‘real’ money accounts.
1.2.1
Physical market participants
Individual product supply chains will be considered in the respective chapter. In general
terms, the commodity will need to be produced, refined, and then transformed into
a product that can be consumed by the end user. Admittedly this general description
does not capture all the different types of commodity supply chains, but the key point
is that the participant will typically have some form of price risk at most points along
the supply chain
In simple terms, producers will be exposed to falling prices, consumers will be
exposed to rising prices, and refiners, processors, and utilities will be exposed to
4
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
margins (e.g. the income generated from selling gasoline less the cost of buying crude
oil). These participants are also faced with a variety of other risks which include:
▪ Credit, i.e. the unwillingness or inability of a customer to pay their debts.
▪ Logistical risks surrounding the movement of the commodity.
▪ Sourcing the right quality of commodity.
▪ Being able to finance day-to-day operations.
1.2.2
Price reporting agencies (PRAs)
One of the problems faced by various commodity markets over the years is one of price
discovery. How does a market participant know if they are achieving fair market value?
Consider the following quote from a market participant in 2011 with respect to the metal
Rhodium, which was about to be used in the creation of an exchange traded fund aimed
at the retail market:
‘With no futures benchmark . . . all the spot price transparency of molasses . . .
and a risk reward with which only a supremely knowledgeable professional or
those wet behind the ears would be comfortable . . . guess the target audience?’
(Financial Times, 2011)
Since commodities are heterogeneous products, establishing a fair price has always
been a challenge for market participants and the main conventions used either involve
exchange traded prices (where available) or index values determined by PRAs. IOSCO
(2012) defines a PRA as:
‘Publishers and information providers who report prices transacted in physical
and some derivative markets and give informed assessment of price levels at
distinct points in time’.
They defined a crude oil assessment as:
‘The process of applying a methodology and/or judgement to market data and
other information to reach a conclusion about the price of oil’.
In response to IOSCO, one of the PRAs, Platts (2012) described their activities in
relation to crude oil as follows:
‘Platts publishes assessments of spot prices for crude oil and refined products in
various geographic regions based on a range of factual inputs including information on individual transactions supplied by market participants . . . Given
the heterogeneous nature of the underlying transactions (in terms of trading
parties, product quality, location, timing, delivery terms and other factors), the
analysis conducted by Platts in determining its published price assessments
is essentially qualitative, albeit based on a range of quantitative and factual
inputs’.
5
Fundamentals of Commodities and Derivatives
Price indexes can be used as the basis for settling commercial supply contracts (as
could futures prices) or, in some cases used to determine the value of cash-settled futures
transactions.
1.2.3
Investment banks
The services offered by investment banks will vary according to the business model that
they adopt. Some banks may consider themselves to be a ‘full service’ entity, offering
the ability to deal not just in derivatives, but to become counterparty to a physical trade.
Other banks may offer more limited services, such as having a derivative service without
the capability to execute physical transactions.
A ‘full service’ bank will be able to offer several solutions to physical participants
with some hypothetical examples shown in Table 1.1.
1.2.4
Commodity trading houses
The commodity trading house Glencore Xstrata describe themselves as follows
(Glencore, 2011): ‘(the company) is a leading integrated producer and marketer of
commodities, with worldwide activities in the marketing of metals and minerals,
energy products and agricultural products and the production, refinement, processing,
storage and transport of these products. Glencore operates globally, marketing and
distributing physical commodities sourced from third party producers and own
production to industrial consumers’. Traditionally, commodity trading houses would
have simply bought commodities from producers and then sold them to consumers.
However, the definition presented by Glencore Xstrata suggests that over time these
entities have evolved to own and operate significant parts of various commodity supply
chains. So, the notion of one company being fully integrated along a supply chain is
no longer the norm. Indeed, many of the investment banking services highlighted in
Table 1.1 could conceivably be offered by trading houses.
A report by the Financial Times (2013) highlighted the extent of trading house
involvement in the market:
TABLE 1.1 Examples of services that could be provided by banks to facilitate commodity
production and consumption.
Sector
Problem
Solution
Crude oil and refined
products
Natural gas
Inventory is working capital
intensive
Cold weather creates increased
demand, but there are delivery
constraints with existing
pipeline infrastructure
Consumers seek favourable
payment terms
Bank agrees to own inventory
Base metals
Banks buy pipelines or own
storage facilities
Banks provide finance along the
logistics chain or act as an
intermediary between
consumers and producers
6
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Those trading oil handled more than 15m barrels of oil a day.
▪ The main agricultural trading houses handled about half of the world’s grain and
soybean trade flows.
▪ Two trading houses controlled about 60% of the zinc market.
▪ Relatively unknown companies can dominate smaller niche markets such as coffee.
Their growth was attributed to four main factors:
▪ The economic boom after 2000 in several emerging economies.
▪ A strategic decision to acquire physical assets.
▪ Their ability to exploit price arbitrage opportunities because of their increasing
presence along the supply chain.
▪ Consolidation in the period prior to 2000 which reduced competition.
1.2.5
Hedge funds
There is no single definition of a hedge fund given the wide range of structures and
strategies used in this section of the market. However, they can be defined in terms of
their characteristics:
▪ Investment ‘vehicles’ that pool the proceeds of their investor base.
▪ Access tends only to be available to a limited group of investors.
▪ Proceeds are actively managed.
▪ They aim to generate a return irrespective of underlying market conditions.
They are often associated with aggressive ‘view driven’ strategies and Chapter 3
includes a case study of the commodity hedge fund Amaranth that failed when its strategy in the US natural gas markets resulted in substantial losses.
1.2.6
‘Real money’ accounts
A real money participant is usually classified as an entity that is not able to borrow
money to boost their available investment proceeds. Typically, this could include entities
such as pension funds and insurance companies. Their participation in the commodity
market is primarily for investment purposes. Also within this category it may be possible to include private and commercial banks that are offering commodity investment
products to their retail customers.
1.3
TRADED VERSUS NON-TRADED COMMODITIES
One of the subtle characteristics of commodity markets is the difference between traded
and non-traded commodities. What is the difference?
An interesting case study that illustrates the key differences is the market for iron
ore. Iron ore is used in the production of steel and, combined with steel, represents the
world’s second largest commodity bloc by value (ICE, 2009). Macquarie Bank (2013)
Fundamentals of Commodities and Derivatives
7
points out that prior to 2003 the concept of a spot market for the metal did not exist in
any meaningful sense. At this time, the traditional buyers were Japanese and Korean
steel producers who purchased their metal using annual, fixed price, bilateral contracts
with suppliers based mainly in Brazil and Australia. The annual benchmark price
typically ran from 1 April–31 March in the following year. Emerging new consumers
such as China struggled to purchase the required amount of metal under this market
mechanism as the traditional sources of supply could not keep pace with the extra
demand. This coincided with a new source of supply from India that was able to react
quickly. This led to more ‘one-off’ transactions that resulted in the emergence of a spot
market. At the same time commodities that were inputs to the steel making process,
which already had developed spot markets, became more volatile. This increased the
pressure on iron ore to respond accordingly.
However, the phrase ‘spot’ within the context of commodities can sometimes be
applied ambiguously. For example, in certain markets (e.g. gold), spot transactions will
have a similar maturity to those seen in traditional financial markets (e.g. trade date
plus two good business days). In other instances (e.g. crude oil), delivery is unlikely to
occur in such a short time frame. ‘Spot pricing’ could also indicate that the contract is
for short-term delivery with prices possibly referencing exchange traded futures prices.
The increase in spot transactions meant that price-reporting companies now disseminated information on physical transactions on a more regular basis. One of the
characteristics noted earlier is that commodity markets lack homogeneity and therefore pricing from a single benchmark has become the accepted practice. For iron ore a
popular benchmark that has emerged is iron ore with a grade of 62%1 .
The development of pricing benchmarks is an important step in the development
of a commodity’s tradable status:
▪ They represent a standard reference point, which is based on actual market activity
and is understood by market participants.
▪ Participants can enter commercial contracts or reference financial contracts to a
price that is transparent, representative of the most liquid market, and is determined by a publicly available process.
▪ The benchmark price is the price of the commodity if it is used by many and varied
participants.
▪ Once a benchmark price emerges, market participants can trade different grades
of the commodity as a differential. So, iron ore with a grade of 58% would trade
below the benchmark price, while a 65% grade would trade above the price. These
differentials may also reflect the products’ country of origin and its availability.
▪ The emergence of a benchmark may result in the development of financial markets
(e.g. derivatives) that can facilitate the hedging of underlying exposures.
1
Ore is essentially the rock that is extracted from through the mining process, which is then
refined to extract the desired elements, such as iron. The ores may be classified by the amount of
desired element that they contain. For example, iron ore ‘fines’ (heavy grains) can vary in grade
from 30% to the mid 60%. (ICE, 2009)
8
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The increased reporting of iron ore prices meant that the spot price provided
a reference point for those entities still using the annual benchmark negotiations.
Indeed, because of the financial crisis the spot price of the metal fell below the annual
benchmark. This resulted in several consumers defaulting on their fixed price agreements in order to take advantage of the lower spot price. Once price assessments started
to appear daily, which reported a standard grade of iron ore, financial products began
to emerge. Exchange traded iron ore swaps were the first derivative that referenced
this product. Thereafter, a forward curve for iron ore started to form, which was later
boosted by the emergence of over-the-counter swaps and iron ore futures.
1.4
FORWARD CONTRACTS
A forward contract will fix the price today for delivery of an asset in the future. Gold
sold for spot value will involve the exchange of cash for the metal in two days’ time.
However, if the transaction required the delivery in, say, one month’s time, it would be
classified as a forward transaction. Forward contracts are negotiated bilaterally between
the buyer and seller and are often characterized as being ‘over-the-counter’ (OTC).
The forward transaction represents a contractual commitment and so, if gold is
bought forward at USD 1,430.00 an ounce, but the price of gold in the spot market is
only USD 1,420.00 at the point of delivery, one cannot walk away from the forward
contract and try to buy it in the underlying market. However, it is possible for both
parties to mutually agree to terminate the contract early. This could be achieved by
agreeing upon a ‘break’ amount, which would reflect the current economic value of the
contract. Typically, this is done using a process that is referred to generically as ‘marking
to market’. An easier way to understand the issue is to use the concept of an exit price.
This is typically taken to be the amount for which an asset could be sold, or a liability
settled in an ‘arm’s length’ transaction.
A variation on the standard contract is a floating forward. In this type of transaction,
a market participant commits to buy or sell the underlying at a future date, but the
applicable price is only set at the point of delivery. The final price that is agreed upon
may be based on some pre-agreed formula. For example, the price could be the average
of daily spot prices in the month prior to settlement.
1.5
FUTURES
A futures contract is traded on an organised exchange, with the CME Group being one
example. Economically, a future achieves the same result as a forward by offering price
certainty for a period in the future. However, the key difference between the contracts
is in how they are traded. The contracts are uniform in their trading size, which is set by
the exchange. For example, the main features of the contract specification for the gold
future appear in Table 1.2.
Fundamentals of Commodities and Derivatives
9
TABLE 1.2 Gold futures contract specification.
Trading unit
100 troy ounces
Price quotation
Trading months
US dollars and cents per troy ounce
Trading is conducted for delivery during the current calendar
month; the next two calendar months; any February, April,
August, and October falling within a 23-month period; and
any June and December falling within a 72-month period
beginning with the current month.
USD 0.10 (10c) per troy ounce (USD 10.00 per contract)
Trading terminates at the close of business on the third to last
business day of the maturing delivery month.
Deliverable
The first delivery day is the first business day of the delivery
month. The last delivery day is the last business day of the
delivery month.
Margins are required for open futures positions.
Minimum price fluctuation
Last trading day
Settlement method
Delivery period
Margin requirements
Source: CME Group
Traditionally, there are some fundamental differences between commodity and
financial products traded on an exchange basis. Historically, one of the key differences
is that futures require collateral to be deposited when a trade is executed (known as
initial margin). As a rule of thumb, the initial margin will be about 5% of the market
value of the contract. Although different exchanges will work in different ways, the
remittance of profits and losses may take place on an ongoing basis (variation margin)
rather than at the maturity of the contract. However, financial markets have evolved
such that OTC forward contracts will now have very similar margining requirements
to futures contracts.
Another difference between forwards and futures relates to the grade and quality
specification. If one is delivering a currency, the underlying asset is homogenous – a
dollar is always a dollar. However, because metals vary in shape, grade, and quality, it
is important to ensure an element of standardisation so the buyer knows what they are
receiving. Some of the criteria that CME Group apply include:
▪ The seller must deliver 100 troy ounces (+/-5%) of refined gold.
▪ The gold must be of a fineness of no less than 0.995.
▪ It must be cast either in one bar or three one-kilogram bars.
▪ The gold must bear a serial number and identifying stamp of a refiner approved
and listed by the Exchange.
Anecdotal estimates suggest that the vast majority (ca. 95%) of futures contracts are
terminated prior to their expiry date. This is perhaps a reflection that most participants
will use the instruments for risk management purposes rather than as a source of supply.
10
1.6
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
SWAPS
In a swap transaction two parties agree to exchange cash flows, whose size are based
on different price indices. Typically, this is represented as an agreed fixed rate against
a variable or floating rate. Swaps are traded on an agreed notional amount, which is
not exchanged but establishes the magnitude of the fixed and floating cash flows. Swap
contracts are typically of longer-term maturity (i.e. greater than one year) but the exact
terms of the contract will be open to negotiation. For example, in many base metal
markets a swap transaction is often nothing more than a single period forward. This is
because the forward transaction may be cash settled which would involve the payment
of the agreed forward price against the spot price at expiry.
The exact form may vary between markets, with the following merely a sample of
how they may be applied in a variety of different commodity markets.
▪ Gold – Pay fixed lease rate vs. receive variable lease rate
▪ Base metals – pay fixed aluminium price vs. receive average price of near dated
aluminium future
▪ Oil – pay fixed West Texas Intermediate (WTI) price vs. receive average price of near
dated WTI future
Swaps will usually be spot starting and so become effective two days after they are
traded. However, it is also possible for the swap to become effective sometime in the
future – a forward starting swap. The frequency with which the cash flows are settled
is open to negotiation but they could vary in tenor between 1–12 months. Where the
payments coincide, there is a net settlement between the two parties. One of the features
of commodity swaps not shared by financial swaps is the use of an average rate for the
floating leg. This is because many of the underlying exposures that commodity swaps
are designed to hedge will be based on some form of average price.
The motivation for entering a swap will differ between counterparties. For a
corporate entity one of their main concerns is risk transference. Take a company that
purchases a particular commodity at the market price at regular periods in the future.
To offset the risk that the underlying price may increase, they would receive a cash
flow under the swap based on movements in the market price of the commodity and
pay a fixed rate. If the counterparty to the transaction were an investment bank, the
latter would now have the original exposure faced by the corporate. The investment
bank would be receiving fixed and paying a variable rate, leaving them exposed to
a rise in the price of the underlying commodity. In turn, the investment bank will
attempt to mitigate this exposure by entering some form of offsetting transaction. The
simplest form of this offsetting deal would be an equal and opposite swap transaction.
To ensure that the bank makes some money from this second transaction, the amount
they receive from the corporate should offset the amount paid to the offsetting swap
counterparty.
11
Fundamentals of Commodities and Derivatives
Swaps are typically traded on a bid-offer spread basis. From a market maker’s perspective (that is the institution giving the quote) the trades are quoted as follows:
Bid
Offer
Pay fixed
Receive floating
Buy
Long
Receive fixed
Pay floating
Sell
Short
Although the terms buy and sell are often used in swap quotes, the actual meanings
are often confusing to anyone looking at the market for the first time. In the author’s
opinion, the most unambiguous way to trade these instruments is to state who is the
payer and who is the receiver of the fixed rate. The convention in all swap markets is
that the buyer is receiving a stream of variable cash flows for which the price is a single
fixed rate. Selling a swap requires the delivery of a stream of floating cash flows for
which the compensation is a single fixed rate.
1.7
OPTIONS
A forward contract offers price certainty to both counterparties. However, the buyer
of a forward is locked into paying a fixed price for a particular commodity. This transaction will be valuable if the price of the commodity subsequently rises, but will be
unprofitable in the event of a fall in price. An option contract offers the best of both
worlds. It will offer the buyer of the contract protection if the price of the underlying moves against them, but allows them to walk away from the deal if the underlying
price moves in their favour.
This leads to the definition of an option as the right, but not the obligation, to either
buy or sell an underlying commodity sometime in the future at a price agreed upon
today. An option that allows the holder to buy the underlying asset is referred to as a
call. Having the right to sell something is referred to as a put. The price at which the two
parties will trade if the option is exercised is referred to as either the strike price or the
exercise price. The strike can be set at any level and is negotiated between the option
buyer and seller.
Options may be either physically settled (that is, the commodity is delivered/
received) or cash settled. The process of cash settlement removes the need to make or
take delivery of the underlying asset, but retains the economics of a physically settled
option. Cash settlement involves the seller paying the buyer the difference between
the strike and the spot price at the point of exercise. The payoff for a cash-settled call
option is:
MAX (underlying price − strike price, 0)
12
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Where: MAX means ‘the maximum of’
The payoff for a cash-settled put option is:
MAX (strike price − underlying price, 0)
Options come in a variety of styles relating to when the holder can exercise their
right. A European style option allows the holder to exercise the option only on the final
maturity date. An American style option allows the holder to exercise the option at any
time prior to final maturity. A Bermudan option allows the holder to exercise the option
on a pre-agreed set of dates prior to maturity.
An option that is in-the-money (ITM) describes a situation where it would be more
advantageous to trade at the strike rather than the underlying market price. Take for
example an option to buy gold at USD 1,400 an ounce when the current spot price is say,
USD 1,425. The option to buy at the strike is more attractive than the current market
price. Where the option is out-of-the-money (OTM), the strike is less attractive than
the market price. If the same option had a strike rate of USD 1,430, the higher strike
makes the option less attractive than buying the underlying at a price of USD 1,420.
Finally, an option where the strike is equal to the current market price is referred to as
an at-the-money (ATM).
Since options confer rights to the holder, a premium is payable by the buyer. Typically, this is paid upfront, but certain option structures are constructed to be zero premium or may involve deferment of the premium to a later date. Premiums on options
are quoted in the same units as the underlying asset. So, since physical gold is quoted
in dollars per troy ounce, the premium will be quoted in the same manner.
Many of the derivatives strategies based on options that are discussed within the
text are illustrated based on the value of the option at maturity. These are illustrated as
follows:
In the upper left-hand of Figure 1.2, the purchase of a call option is illustrated.
If at expiry the option market price is lower than the strike, the option is not exercised,
and the buyer loses the premium paid. If the underlying price is higher than the strike
price, the option is exercised and the buyer receives the underlying asset (or its cash
equivalent), which is now worth more in the underlying market than the price paid (i.e.
the strike price). This profit profile is shown to the right of the strike price. On the other
side of the transaction there is the seller of a call option (top right quadrant of Figure 1.2).
The profit and loss profile of the seller must be the mirror of that of the buyer. So,
in the case of the call option, the seller will keep the premium if the underlying price
is less than the strike price, but will face increasing losses as the underlying market
price rises.
The purchase of a put option is illustrated in the bottom left quadrant of Figure 1.2.
Since this type of option allows the buyer to sell the underlying asset at a given strike
price, this option will only be exercised if the underlying price falls. If the underlying price rises, the buyer loses the premium paid. Again, the selling profile for the put
is the mirror image of that faced by the buyer. That is, if the underlying price falls, the
seller will be faced with increasing losses. However, if the market price rises, the seller
will keep the premium.
13
Fundamentals of Commodities and Derivatives
Buy Call
Sell Call
Profit
Profit
Strike
Strike
Underlying
price
Loss
Underlying
price
Loss
Buy Put
Sell Put
Profit
Profit
Strike
Underlying
price
Loss
Underlying
price
Loss
FIGURE 1.2 Profit and loss profiles for options at expiry.
Options arguably offer great flexibility to the end user. Depending on their motivation, it can be argued that option usage are categorised in four different ways. Firstly,
they can be used to take a directional exposure to the underlying market. For example,
if a user thought the underlying price of gold might increase, they could buy physical
gold, a future, or a call option. Buying physical gold requires the outlay of proceeds,
which may need to be borrowed. Buying a future reduces the initial outlay of the physical but will incur a loss if the future’s price falls. Buying a call option involves some
outlay in the form of premium but allows for full price participation above the strike
and limited downsides if the price falls. The second usage for options is an asset class
in its own right. Options possess a unique feature in implied volatility, and this can be
isolated and traded. The focus of this type of strategy is how the option behaves prior to
its maturity. The third motivation, which is particularly relevant to the corporate world,
is as a hedging vehicle that allows a different profile than that of the forward. With a
little imagination it is possible to structure solutions that will offer differing degrees
of protection against the ability to profit from a favourable movement in the underlying price. The final motivation is options as a source of outperformance. For example,
if an end user owns the underlying asset (e.g. central bank holdings of gold) they can
use options to exceed some performance benchmark such as money market deposit
rates. From a hedger’s perspective, options could be used to outperform an ordinary
forward rate.
14
1.8
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
EXOTIC OPTIONS
Exotic options are a separate class of options where the profit and losses upon exercise
do not correspond to the plain vanilla American and European styles. Although there
is a proliferation of different types of exotic options, it is worth introducing some of the
key building blocks, which feature prominently in derivative structures.
1.8.1
Binary options
A binary option (sometimes referred to as a digital) is very similar to a simple bet. The
buyer pays a premium and agrees to receive a fixed return. Very often the strike rate on
the digital is referred to (somewhat confusingly) as a barrier. With a European style call
option, the holder will deliver the strike price to the seller and receive a fixed amount of
gold. However, the value of the gold will depend on where the value of gold is trading in
the spot market upon exercise. With a binary option the buyer will receive a fixed sum
of money if the option is exercised, irrespective of the final spot level.
An example of a long digital call and a long digital put is shown in Figure 1.3.
1.8.2
Barrier options
The purchaser of a barrier option will either:
1. Start with a conventional ‘plain vanilla’ option that could subsequently be cancelled
prior to maturity (known as a knock out), or,
2. Start with nothing and be granted a conventional option prior to the maturity of
the transaction (known as a knock in).
Profit
Profit
Underlying price
Underlying price
Loss
Digital Call
FIGURE 1.3 Digital calls and puts.
Loss
Digital Put
15
Fundamentals of Commodities and Derivatives
The cancellation or granting of the option will be conditional upon the spot level in
the underlying market reaching a certain level, referred to either as a barrier or trigger
level.
The position of the barrier could either be placed in the out-of-the money region
or in the in-the-money region. This will be above or below the current spot price, as we
will show below. The former are referred to as standard barriers with the latter known
as reverse barriers. This could result in what may initially seem like a bewildering array
of possibilities. Figure 1.4 summarises the concepts.
To illustrate the concept further let us return to the option example illustrated
earlier and concentrate on analysing a call option. We will assume the option is
out-of-the-money and the current market conditions exist:
Spot
Strike
Maturity
USD 1,425
USD 1,430
Three months
The purchaser of a standard knock in barrier option would be granted a European
style option if spot hit a certain trigger. Since it is a standard barrier option, the trigger
must be placed in the out-of-the-money region so it would be set at say, USD 1,420. Consequently, spot must reach USD 1,420 or below before the option is activated (knocked
in), hence the name a down and in. If the purchaser started with a standard barrier call
option with the trigger at USD 1,420, it would be a down and out. That is, if spot were
to fall to USD 1,420 or lower the option contract would be cancelled. A reverse knock in
call option would have the barrier placed in the money at, say, USD 1,435. A purchaser
of such an option would have a contract that would grant a call option with a strike of
USD 1,430 if spot hits USD 1,435. The final example would be a reverse knock out call
option, with the trigger again set at USD 1,435. Here the purchaser starts the transaction
with a regular call option, which would be cancelled if spot reached USD 1,435 – an up
and out contract.
Barrier options
Standard
barriers
Knock in
options
Call option
‘down and in’
Put option
‘up and in’
Reverse
barriers
Knock out
option
Call option
‘down and out’
Put option
‘up and out’
FIGURE 1.4 Taxonomy of barrier options.
Knock in
option
Call option
‘up and in’
Put option
‘down and in’
Knock out
option
Call option
‘up and out’
Put option
‘down and out’
16
1.8.3
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Spread options
A spread option pays off based on the difference between the price of two underlying
assets, relative to a pre-agreed strike. A call option will pay off if the spread is greater
than the strike, while a put option will pay off if the spread is lower than the strike.
Call payoff = MAX (Price of asset 1 − price of asset 2 − strike, 0)
Put payoff = MAX (Strike − price of asset 1 + price of asset 2, 0)
Spread options are relatively more popular in commodities than in traditional
financial assets. For example, a participant may wish to hedge or take exposure
between:
▪ The cost of an input and the revenue earned from an output (e.g. cost of crude oil
and revenue from gasoline).
▪ The prices between a commodity traded at two different locations.
▪ The price of a single commodity future for two different delivery dates.
These examples will be discussed in detail in their respective chapters. From a valuation perspective, these options require an additional valuation input and that is the
correlation that exists between the two asset prices. To illustrate this, consider a spread
option structured as a call, which will pay off if the spread between two asset prices
increases beyond a pre-agreed strike. If the two asset prices are negatively correlated,
an increase in asset price 1 will be associated with a decrease in asset price 2. This relationship increases the probability that the option will end up more deeply in the money
and as such the seller will charge a higher price.
1.8.4
Average rate options
Anecdotally, average rate options (‘avros’) are the most common type of commodity
option. Regular or ‘vanilla’ option payoffs will reference a single underlying asset price
at maturity, with the payoff being made relative to the strike. However, an average rate
option will pay out based on an average of prices covering a pre-agreed period prior to
maturity.
The expiry payoffs of average rate options are:
Call option = MAX (average underlying price − strike price, 0)
Put option = MAX (strike price − average underlying price, 0)
One of the reasons for the popularity of such contracts in commodities is that it
reflects the way in which physical supply contracts are structured. A refiner who agrees
to buy crude oil from a producer will avoid paying a potentially high price if the terms
of the deal reference market prices on a single date. Although averaging prices will
dampen the effect of volatile price movements it also means that the refiner will not
be able to benefit if the price of the commodity suddenly falls. A common feature of
Fundamentals of Commodities and Derivatives
17
commodity derivatives is that the payoff on the instrument should match that of the
commercial contract to avoid a cash flow mismatch.
One of the characteristics of an average rate option is that they will show a premium
reduction over the equivalent European option; it is not really a cheap alternative to a
vanilla option. As an example, the cost of a six-month call option on a crude oil future
with an at-the-money strike price of USD 50.00 and an implied volatility of 30%. The
cost of a vanilla option on this asset is USD 4.20/barrel, but if the terms of the contract are altered such that the final settlement price is an average of the futures prices
over the last three months of the transaction, the cost of the option falls to USD 3.96.
This is because the implied volatility of the average rate option references an average
price series which will be lower than a non-averaged equivalent. Another way of thinking about the problem is that since the premium of an option is a function of payoff
received by the buyer, a payoff based on an average price series will always be lower
than a non-averaged equivalent.
CHAPTER
2
Derivative Valuation
T
he individual demand and supply dynamics for each commodity will be analysed
within the respective chapter. This section considers some of the bigger picture
issues that will influence prices.
2.1
ASSET CHARACTERISTICS
Greer (1997) argues that there are three different types of asset class.
▪ Capital assets – examples include bonds, which pay an investor a stream of fixed
cash flows and equities, which would normally pay a stream of dividends. These
types of assets can be valued using discounted cash flow (DCF) techniques. For
example, for bonds, value is determined by present valuing the fixed cash flows.
For equities, techniques such as the dividend discount model can also be used for
valuation purposes.
▪ Consumable/transformable assets – According to Greer: ‘You can consume it. You
can transform it into another asset. It has economic value. But it does not yield an
ongoing stream of value’. Examples of this type of asset would include most commodities such as agriculture, metals, and energy. Greer’s comment is significant:
‘the profound implication of this distinction is consumable/transformable (C/T)
assets . . . cannot be valued using net present value (NPV) analysis. C/T assets must
be valued more often based on the particular supply and demand characteristics of
their specific market.’
▪ Store of value assets – this type of asset cannot be consumed, nor does it generate income. However, it does have value. One example given by Greer is fine art.
Auctions of fine art at substantial prices are often reported in the press.
2.2
COMMODITY PRICES AND THE ECONOMIC CYCLE
Figure 2.1 illustrates in a stylised manner the different parts of the economic cycle as it
relates to commodities and the likely price reaction. The diagram is based on Macquarie
(2015) reported in Commodities Now (2015).
18
19
Derivative Valuation
Strong demand
push or deficit
market
Pricing to encourage
supply
Prices stable
Demand
destruction /
capex accelerating
Supply constraint /
Capex lagging
Stocks drawing /
limited new supply
Prices stabilise at low level
High stocks /
strong supply
reaction
Price acceleration
Price peaking
Market in balance
/ capex peaking
Strong supply
growth / stocks
building
Prices falling
Severe price decline
FIGURE 2.1 Commodity prices and the economic cycle.
Source: Macquarie
2.3
PRINCIPLES OF COMMODITY VALUATION
Within the context of commodities, analysts who have gravitated from traditional financial assets have tried to apply NPV techniques to value commodities. However, the
application of these models has proved to be problematic because, as Greer points out,
commodities are not bonds and equities. Barclays (2011) argues, ‘Most financial assets
can be understood and valued through a cash flow discounting technique approach.
Commodities are different – they are not capital assets and are subject to unique supply
and demand forces that lead to complex valuation dynamics’. Examples of some of these
dynamics are given below:
Inventories
One subtle fact of commodity price behaviour relates to the balance of supply
that is either above or below ground. Typically for gold there is substantial above
ground inventories and as such, short-term prices are not significantly influenced
by changes in either fundamental demand or supply (e.g. a change in mine supply).
However, for a metal such as copper, the dynamics tend to be different with lower
levels of inventories being common resulting in greater short-term price sensitivity
to changes in fundamental demand and supply.
Seasonality
The demand and supply for certain commodities may only occur at certain
times in the year. So, the demand for heating oil may be greater in winter than
in summer. This would result in higher prices for winter delivery and lower prices
for summer delivery.
20
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Price inelasticity
One of the ways to understand commodity price formation is to appreciate that
short demand and supply are relatively inelastic (Morgan Stanley, 2009). Inelasticity describes how easily prices will respond to changes in demand and supply
factors. Inelastic demand occurs when a change in price does not immediately
impact the demand for a product:
▪ The cost of the commodity is a small component of the final product – if the
price of coffee beans were to increase, the price of a consumer’s daily caffeine
drink may not change significantly.
▪ The commodity is an essential input – planes need jet fuel to fly; gold rings
require gold as an input. For example, during periods of increased energy
prices, airlines have been able to increase fares by adding a fuel surcharge
fee to tickets. It may well be that people decide to economise, but this is an
example of the cost being passed to the consumer.
▪ Product substitution – Another key theme within commodity price formation is the ability to substitute a cheaper alternative commodity for the more
expensive input. This may be difficult with respect to airplanes, which are
configured to burn a specific type of fuel.
▪ New technology – prices of commodities will evolve as new technology is
implemented. However, this factor will have a greater impact in the longer
rather than shorter term.
▪ Product and brand loyalty – the author confesses to being a coffee addict. He
does not care how much his Americano will cost as part of the pleasure of
the coffee shop environment (and the free Wi-Fi).
Inelastic supply occurs when output cannot respond quickly to a change in
price:
▪ There may be a finite supply of the product, e.g. carbon-based energy sources.
▪ There may be a finite amount of a related production input, e.g. land or water
supplies.
▪ There may be a lack of infrastructure within the country of supply that will
make it difficult to increase supply on a short-term basis.
▪ Geopolitical issues may have an adverse impact on future supply.
▪ New technology to increase supply will take time to have an effect.
Marginal cost
Marginal cost describes how the total cost of producing a commodity changes
with each incremental unit. Within the context of commodity valuation, it is sometimes described as ‘cost support’. Theoretically, the marginal cost of producing a
commodity should represent the minimum selling price for a commodity, but this
condition may not always hold in the short term. ‘While costs remain a critical variable in forecasting commodity price, they are more relevant for gauging long-term
pricing; in the short term, there is much less rigid support in bear markets than
many expect.’ (Credit Suisse, 2013). They also make the following points as to why
this situation may arise:
– In the short term, many costs are fixed, and so excess supply may lead to
prices falling below the cost of production.
Derivative Valuation
21
– A producer may be willing to suffer short-term losses in anticipation that
their fortunes may improve rapidly.
Mean reversion
Mean reversion is the tendency for prices to revert to some long-term average
value. Casual empiricism suggests that commodity prices do not continually rise
or fall although they are certainly volatile. However, this is an imprecise concept as
there is no standard definition of what constitutes long-term.
2.4
FORWARD PRICE CURVES
A forward (or futures) price is a value that a participant can fix today that will apply to
some future time. It is a representation of the value in the future attributed by the market
to a commodity at a single point in time based on trading activities. For example, if crude
oil prices have fallen, a refinery may take the opportunity to fix the price for oil to be
delivered in say, six months’ time. Arguably, understanding how commodity forward
curves move is central to understanding market behaviour.
Figure 2.2 illustrates the forward price curve for West Texas Intermediate.
The chart illustrates some interesting features of forward price behaviour:
▪ In some circumstances short-dated forward prices are higher than long-dated prices
and vice versa. These conditions are referred to as backwardation and contango
respectively and are analysed in greater detail in the next section.
▪ Shorter-dated contracts are more volatile than longer-dated contracts.
▪ At low prices, the forward price curve slopes upward. As prices rise, the curve flattens and eventually inverts.
2.4.1
Forward prices – a market in contango
Within the commodities world, there are two ways of describing the state of a forward
market: contango or backwardation. Contango describes a situation where the price for
forward delivery is higher than the price for spot delivery, while backwardation exists
when the forward price is below the spot price. Although both of these states exist in
the pricing of traditional financial products, the role of the underlying physical markets
in commodities is much more important, particularly when demand exceeds supply.
It is very rare for financial products to experience prolonged periods where the principles of ‘no arbitrage valuation’ do not hold. Consider for a moment the world of equity
indices. The fair forward value of an equity index is determined using a technique that
is sometimes referred to casually as ‘cash and carry’ pricing. Suppose that an investment
banking client wishes to buy a basket of equities that exactly matches the composition
of the domestic equity index for delivery in one year’s time. To hedge this exposure,
the bank will immediately buy all of the stocks according to their respective weights in
the index, which is assumed to be trading at a spot value of 100 points. To finance the
purchase of these shares the bank borrows the equivalent of 100 index points at its cost
of funding which is assumed to be 3% p.a. This would result in a funding cost equivalent of three index points. While holding the stocks for one year, it is assumed that the
22
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
120.00
110.00
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00
Spot
2009
1 year
2011
2 year
2013
3 year
4 year
5 year
2014
2015
6 year
2016
FIGURE 2.2 Forward price curves.
Source: CME Group
stocks will pay a dividend equivalent to 1% p.a., i.e. one index point. Since the stocks
are not required for 12 months, they can be lent out to earn some fee income, which is
assumed to be the equivalent of 0.5 index points. This means that the fair value of the
index constituents at maturity will be:
Forward price = spot price + financing cost − dividend income
− securities lending fee
= 100 + 3 − 1 − 0.5
= 101.5m
So, 101.5 index points is the minimum that the bank will charge for this transaction.
It is important to realise that this forward is not a forecast. It is a mechanical breakeven
calculation, not ‘the market’s best guess of where the future spot price will be’.
Now imagine a slightly different scenario, but based on the same observed values.
Suppose a trader sees another bank quoting a 12-month forward on the index of 101.
This would suggest that this forward is trading below ‘fair value’ of 101.5. Ignoring all
transaction costs and bid offer spreads, the trader could profit from this mispricing by
executing the following transaction:
▪ Buy the index forward for delivery in 12 months at an agreed cost of 101 from the
other bank.
▪ Borrow the constituent shares and sell them for spot value raising the equivalent of
100 index points.
Derivative Valuation
23
▪ Place the proceeds of the sale on deposit for 12 months to earn three index points
of interest.
▪ Pay the lender of the shares the one index point of dividend income received during
the year1 . The bank will also have to pay a borrowing fee, which is assumed to be
0.5 index points.
▪ The net result of these trades is that the forward purchase incurs a cost of 101 index
points while the combination of the spot trades results in income of 101.5 (+100
+ 3 − 1 − 0.5). Overall, the bank has been able to lock in an arbitrage profit of 0.5
index point.
Within the equities market such anomalies would be very short-lived given the liquidity of the market and the speed with which trades can be executed.
Although generally speaking commodities are different, the forward valuation of
certain commodities will follow the time value of principles outlined above. One such
example is gold, which is often traded like a financial commodity.
We will use the example of a gold producer who approaches a bank asking for a
price for delivery of gold in, say, six months. The price quoted by the bank is not a guess
and neither is it a forecast of where it thinks the price of gold will be at the time of
delivery. Rather the price quoted will be driven by the cost of hedging the bank’s own
exposure. This illustrates one of the key maxims of derivative pricing – the cost of the
product is driven by the cost of the hedge. This approach to valuation is sometimes
referred to as a ‘cash and carry’ trade.
If the bank does not hedge its price exposure then in six months’ time it will take
delivery of gold at the pre-agreed price and will then be holding an asset whose market
value could be lower (or higher) than the price paid to the producer.
To avoid the risk of a fall in the gold price, the bank executes a series of transactions
on the trade date to mitigate the risk. Since the bank is agreeing to receive a fixed amount
of gold in the future, it sells the same amount in the spot market to another institution,
say, another investment bank. However, the bank has sold a quantity of metal today that
it will not take delivery of for six months. To fulfill this spot commitment, it can borrow
the commodity until it receives the gold from the producer. The gold could be borrowed
from a central bank that would be paid interest at the maturity of the loan. Having sold
the gold spot and borrowed to cover the sale, the bank is now holding dollar proceeds.
Since the bank would be looking to manage its cash balances, these dollars would now
be invested until the producer delivers the gold.
As a result, it is possible at the inception of the forward trade to identify all of the
associated cash flows, allowing the bank to quote a ‘fair value’ of theoretical price that
will ensure no loss at the point of delivery, irrespective of the prevailing price. In this
example the maximum amount the bank will pay the producer cannot exceed:
▪ Proceeds received from the spot sale plus,
▪ The interest received from the dollar deposit less,
▪ The interest paid to the lender of gold.
1
In a securities lending deal market convention dictates that the lender of the share will retain
any dividends paid during the life of the loan.
24
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
A simple example may help illustrate the point. We will assume that the producer
asks for a six-month (182 days) forward price. For simplicity we will calculate the forward price for a single ounce. In the cash market, gold is trading at USD 1,425.40 per
ounce, so the dealer agrees to sell one ounce. In order to complete the spot delivery, he
borrows the same amount from the local central bank for 6 months at a lease rate of
0.11570% per annum. The dollars received from the spot sale are put on deposit for 6
months at a LIBOR to earn, say, 3.39% per annum. The interest cost of borrowing the
metal is USD 0.8333 (spot x lease rate x 182/360) and that USD 24.43 is earned from
the cash deposit (spot sale proceeds x 6-month LIBOR x 182/360). So, the maximum
amount he can afford to pay the producer is USD 1,449.10 (rounded). This is calculated
as spot sale proceeds plus interest on LIBOR deposit minus the borrowing fee (USD
1,425.40 + USD 24.43 − USD 0.8338). This fair value is a breakeven price for the trader
and so may be adjusted to build in an element of profit.
So the forward price is the spot price plus the cost of carrying an underlying hedge.
It is important to note that the shorter the time to maturity, the smaller will be the
differential between the spot and forward price since the hedge is carried for a shorter
period. Indeed, if we were to recalculate the forward price applicable for a single fixed
date in the future on a daily basis, the differential would reduce (all other things being
equal) until the final calculation when spot and forward becomes the same thing and
the two prices will have converged.
The observed forward price of a commodity is kept in line with its fair value by the
possibility of arbitrage similar to the principles outlined in the equity example. In the
previous example where the fair value of gold for six-month delivery was USD 1,449.10,
it is unlikely that the market price would be significantly different. Say that a forward
market price of USD 1,440 was observed. With the fair value the instrument calculated
at USD 1,449.10, the commodity would be described as being ‘cheap to fair value’. In
this situation an arbitrager could:
▪ Buy the commodity forward, paying USD 1,440 upon delivery.
▪ Short the gold in the spot market to earn the spot price of USD 1,425.40.
▪ Invest the cash proceeds at LIBOR earning 3.39% for six months to earn USD 24.43.
▪ Borrow gold in the lease market to fulfill the short spot sale paying a six-month
lease rate, which equates to a cash amount of USD 0.8333.
▪ Repay the gold borrowing upon receipt of the metal under the terms of the forward
contract.
The arbitrager would end up with a net profit of USD 9.10 the difference between
the theoretical value of the forward contract (USD 1,449.10) and the market value of the
forward (USD 1,440).
2.4.2
Forward prices – a market in backwardation
Backwardation describes a situation where the prices for shorter-dated forward contracts are higher than those of longer-dated forward contracts. Forward pricing theory
dictates that the market maker quotes a forward price such that all expenses incurred, as
well as any income benefits they may have derived from carrying an underlying hedge,
Derivative Valuation
25
are passed on to the customer. With respect to gold, storage on an unallocated basis is
not considered to be a significant expense, and is therefore traditionally not included in
forward pricing considerations. However, with base metals and energy products, storage
and insurance costs are included. Based on conventional time value of money principles, this should give a forward price relationship of:
Forward price = Spot price + LIBOR + storage and insurance costs
From this relationship it follows that the fair value of a forward contract should
always be greater than the spot value. However, many commodity markets experience
conditions of backwardation (e.g. some base metals, crude oil), which suggests that the
previous equation is incorrectly specified. To explain this apparent anomaly some analysts adjust the equation to include a term referred to as the ‘convenience yield’. So now
the forward price equation reads:
Forward price = Spot price + LIBOR + warehousing and insurance costs
− convenience yield
The convenience yield can be defined as the “flow of services and benefits that
accrues to an owner of a physical commodity, but not to an owner of a contract for
future delivery of the commodity. This can come in the form of having a secure supply
of raw materials and hence, eliminating the costs associated with stock outs.”
The magnitude of the convenience yield will vary according to the physical balance
of demand and for the underlying commodity. If the commodity is in very short supply,
its value will rise, moving towards zero in ‘normal’ supply conditions. If, however, the
minimum value of the convenience yield is zero, there is no maximum as its value is
driven by the consumer’s need to obtain the physical commodity immediately.
Initially, the author found the concept of convenience yield as particularly strange.
Over the years anecdotal conversations with market practitioners have suggested one
common theme: In the real world no one uses convenience yield. Some of the anecdotal
comments made to the author include:
▪ There is no such thing as convenience yield. Commodities yield nothing: they do
not carry coupons and they do not pay dividends.
▪ It is a financial mathematician’s tool to try and describe a market behaviour that
they never witness in traditional financial products and so cannot explain.
▪ It is just a number that makes the traditional forward pricing formula work.
In the case of a backwardated market the forward market price is lower than the
spot price suggesting that the contract is mispriced or trading ‘cheap to fair value’. If this
were the case a speculator should be able to buy the cheap forward contract, sell it for
spot value and hold the combined position to maturity to realise a profit. This strategy,
outlined in the previous section covering contango markets, would allow the arbitrager
to earn the difference between the mispriced forward and its theoretical value. The reason this cannot happen and also why the market will remain in a prolonged state of
26
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
backwardation is that when selling the contract in the spot market, the participant will
need to obtain the commodity to fulfill the short commitment. Since the availability of
the commodity in a backwardated market is by definition very scarce, supplies for spot
delivery are simply not available. Hence this apparent mispricing will persist for prolonged periods, as there is no mechanism to exploit the potential arbitrage, i.e. there is
no deep and liquid market to borrow and lend certain commodities.
An alternative approach that avoids the use of discounting cash flows is suggested
by Greer (2005). He argues that on average a producer tends to have larger operations,
greater inventories, and higher fixed costs than a consumer and so needs greater price
risk protection. Consider the price risk faced by a producer of crude oil looking to sell in
six months’ time. The producer’s current marginal cost is assumed to be USD 45.00, so
this would need to be the minimum he would want to receive in order to break even. He
knows from experience that oil prices can be volatile and difficult to predict. Nonetheless, he has an expectation that the price of crude oil will be USD 50.00 at that time. The
producer approaches a commodities trading house to see if they would be interested
in a forward transaction. The trader happens to share the producer’s opinion on the
expected future spot price and so agrees to a trade. However, he is unwilling to pay USD
50.00 and offers USD 49.00. This is not unreasonable. If the price does evolve as per the
market, the trader will not make any profit. He will buy it from me, the producer, at USD
50.00 and would sell it in the open market at the same price making zero return. The
USD 1.00 ‘discount’ that he suggests is his potential return and is sometimes referred to
as a ‘risk premium’. It is a satisfactory deal for the producer as they have covered at least
their marginal cost of production. This suggests an alternative way of deriving forward
prices (Barclays, 2011):
Forward price = Spot price + expected future spot price
appreciation∕depreciation + risk premium
However, it is instructive to consider the convenience yield as a proxy for the level
of inventories and the availability of storage.
To illustrate this principle, the returns an investor could earn from a commodity
index position that references a portfolio of futures is:
Total returns = spot return + roll yield + collateral yield
(2.1)
The traditional time value of money relationship between the spot and forward
price can be expressed as:
Forward price = spot price + interest rate − (convenience yield − storage)
(2.2)
Rearranging 2.2 to solve for convenience yield gives:
Convenience yield = (Spot − forward) + storage cost + interest cost
(2.3)
27
Derivative Valuation
Within the context of commodity index investing the difference between a spot and
forward price is the roll yield2 .
This results in a formula that allows for the estimation of the convenience yield
based on observable values:
Convenience yield = Roll yield + storage cost + interest cost
(2.4)
Using market prices and industry estimates for storage, Lewis (2005) calculates the
average convenience yield for a range of commodities for the period 1989 to 2004. These
values range from 35.88% for crude oil to -0.84% for gold. He then compares these values to the average number of days of above ground inventories (for example 20 for
crude oil and 16,500 for gold). He argues that convenience yields tend to increase as
the level of inventories decline. In some respects, this is logical; a shortage of supply
in the short-term will cause short-term prices to increase. He describes convenience
yield as an indication of ‘market precariousness’. He goes on to show that there is a
positive relationship between convenience yield and volatility, for those markets that
suffer from lower levels of inventory and hence display a higher convenience yield have
the highest levels of volatility. Equally, products that have high levels of inventory, which
is associated with lower convenience yields, have a lower volatility.
An interesting alternative is offered by Barclays (2008). They suggest that taking
the robot or toddler approach are the two ways to view price determination. In the robot
approach:
The view of prices as being the deterministic output of a calculating device
might just be applicable in a market with close to perfect information, where
the flow of data was rapid and reliable and where the underlying relationships
between fundamentals and prices were well known and stable. Those conditions could just happen in some markets. However, the oil market does not fit
that template, nor is it ever likely to.
They suggest however, that the toddler approach is a more representative view:
[The market] behaves something like a toddler, constantly in search of defining
where the boundaries of behaviour should be, and then constantly pushing
towards those boundaries until it finds them and gets a reaction. Unless there
is something acting in loco parentis that dictates what those boundaries are in
advance, (i.e. normally OPEC), the market will tend to find them for itself.
2.4.3
Interpreting forward curves
Prices for future delivery as represented by the forward curve will experience either
contango (forward prices higher than spot prices) as well as backwardation (forward
prices lower than spot prices).
2
This is covered in Chapter 14.
28
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Although something of a generalisation, commodity sectors such as crude oil and
base metals are more prone to backwardation. This contrasts with precious metals,
which mostly experience contango. One explanation of this is the availability of the
commodity above ground. In the gold market a large amount of gold is held above
ground, and storage costs are relatively small. However, with crude oil it is relatively
expensive to store the commodity (about 20% of the cost of a barrel), which discourages
holding the commodity. As a result, an increase in demand may not be met from existing stocks, as it will need to be extracted and refined. Given the time lag, spot prices rise
relative to forward prices and may push the market into backwardation.
Backwardated markets are characterised by:
▪ A scarcity of the commodity.
▪ Low inventories.
▪ Volatile prices because of low inventories.
▪ A strongly rising price.
▪ A “fear” premium.
Contango markets are characterised by:
▪ A glut of the commodity.
▪ High levels of inventory.
▪ Relative price stability.
▪ General price weakness.
▪ A “complacency” discount.
2.4.4
Commodity arbitrage
Within commodity markets it is perhaps possible to differentiate between three types
of arbitrage:
▪ Geographic
▪ Time
▪ Technical or quality
Geographic arbitrage
One popular example of cross-market arbitrage exists in the metals market between
the London Metal Exchange (LME) and the Shanghai Futures Exchange (SHFE). If the
metal is deliverable to satisfy a futures commitment in either of the markets, then it
may be possible to lock in a profit by buying in one location and selling into the other.
In one sense, this profit opportunity should not happen because if the metal is fungible
between the exchanges, any price differential should be small. However, markets are not
perfect, and as such, differences exist due to demand and supply dynamics, tax regimes,
exchange rates, freight costs, interest rates, and premiums and discounts.
Derivative Valuation
29
The LME (2020) set out the arbitrage formula:
[(LME + ∕ − ADJ + CIF) ∗ USD∕RMB ∗ (1 + VAT) ∗ (1 + TARIFF) + ADMIN]
− SHFE = Arbitrage level
Where:
LME = LME metal price (i.e. three-month future)
ADJ = contango / backwardation adjustment to align the expiry dates of
the LME and SHFE expiration
CIF = any physical premium associated with metal as well as all transport
costs to China
USD / RMB = the exchange rate expressed as the amount of RMB per 1 USD
VAT = value added tax
TARIFF = any import or export taxation
ADMIN = administrative charges
SHFE = equivalent expiration month price quoted on Shanghai Futures
Exchange
So, looking at the arbitrage formula, everything within the square brackets would
represent the ‘all in’ cost of exporting the metal to China. Subtracting the price of the
equivalent contract in Shanghai would return the arbitrage amount. So, the arbitrage
would work if the price of metal for delivery on the SHFE exceeds the all-in cost of
delivering it from the LME. It is also possible to reverse the direction of this arbitrage
by buying on the SHFE and selling into the LME.
Time arbitrage
This type of arbitrage seeks to exploit the shape of the forward curve and would yield
a profit if the forward price were trading away from the price implied by a ‘cash and
carry’ trade. Suppose that aluminium for ‘cash’ (i.e. immediate) delivery is trading at
USD 2,000 per tonne. A trader believes she can store the metal for a week at a cost
of, say, USD 3.00 and the cost of financing the position is, say, USD 0.50. Using cash
and carry principles this implies a ‘fair value’ forward price of USD 2,003.50. However, she notices that the one-week forward is trading at USD 2,005. She executes the
following trade:
▪ She buys the metal for cash delivery at a price of USD 2,000 with the intention of
selling it in a week’s time at the prevailing spot price.
▪ She sells a one-week forward at a price of USD 2,005, which will be closed out
shortly before expiry.
30
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
At the end of the period consider the profit or loss on the position in the following
two scenarios:
Cash price has increased to USD 2,010
The physical position is sold back into the market resulting in a gross profit of USD
10.00 (USD 2,010 − USD 2,000). After substituting carry costs of USD 3.50, the profits
are USD 6.50. It is assumed that the futures position is closed at the prevailing cash price
resulting in a loss of USD 5.00 (USD 2,005 − USD 2,010). The net profit is USD 1.50.
Cash price has fallen to USD 1,990
The physical position is sold back into the market resulting in a gross loss of USD 10.00
and so the net loss incorporating the carry increases this to USD 13.50. Since the price
of the metal has fallen, the short futures position will show a profit of USD 15.00 (USD
2,005 − USD 1,990) resulting in a net profit of USD 1.50.
Note that on both occasions the net profit is USD 1.50, which is equal to the difference between the initial ‘cash and carry’ implied forward price of USD 2,003.50 and the
observed forward price of USD 2,005.
Although some readers may be tempted to suggest that this example is somewhat
theoretical, consider the following: In February 2010, a newspaper report first suggested
that between 75–90% of the world’s physical aluminium stocks were tied up in financial
arbitrage deals designed to exploit the difference between spot and forward prices.
In aluminium, physical stock levels on the London Metal Exchange had quadruped
over an 18-month period to about 4.5 m tonnes, largely because of excess supply relative
to actual demand. However, estimates suggested the amount of aluminium available for
consumption was roughly half that amount.
At the time, the aluminium price was steeply in contango so longer-dated forward
prices were higher than shorter-dated prices. The steepness of the forward curve
allowed the traders to exploit a lucrative arbitrage opportunity. Using 2010 prices, the
profitability of the trade could be calculated as follows:
▪ Buy the metal for spot delivery at USD 2,076.
▪ Finance the purchase of the metal for 92 days at 0.5% (i.e. 3M LIBOR + 0.25%).
▪ Warehousing costs of USD 0.15 per tonne per day.
▪ Sell the metal for three-month forward delivery at USD 2,108.
The difference between the spot cost and forward income is USD 32.00 a tonne. The
financing cost is USD 2.65 and the warehousing amounts to USD 13.80. This resulted
in ‘carry costs’ of USD 16.45 and so generated a net profit of USD 15.55 per tonne or an
annualised return of about 3%.
It would be logical to suggest that profits on such a lucrative trade should disappear
rapidly, but this profitability was consistently reported for about the next five years.
Theory suggests that through demand and supply dynamics, more and more spot
purchases and forward sales would erode the profitability. For the arbitrage to continue,
Derivative Valuation
31
it would require equal and opposite pressure from other market participants. It would
be reasonable to suggest that spot prices would be kept low if producers had excess
inventory to sell, and that forward prices would remain high if consumers were still
buying on a forward basis. Traders engaging in these arbitrage deals were then being
‘rewarded’ for intermediating between the different transactional maturities of the
physical counterparties.
It is not only in metal markets that such arbitrage exists. In December 2008, BP
decided to exploit the shape of the forward crude oil curve by booking supertankers to
store oil at sea. BP booked the Eagle Vienna, a vessel that can store two million barrels
of oil, which is about 10% of America’s daily demand. It was filled with Brent and transported for storage in the Gulf of Mexico for later release to the US market. Although the
company would neither confirm nor deny it, some market participants believed the oil
had been sold on a forward basis. In December 2008, the spot price of crude was quoted
as USD 47.00/barrel while the 12-month forward price was USD 60.00/barrel, so it is
possible to illustrate the economics of the trade using these values.
At the time other market participants such as Royal Dutch Shell and Koch (a petrochemical company) had done the same, leading to suggestions that about 10 million
barrels of oil were being stored in a similar fashion. Estimates suggested that each barrel
would cost about USD 6.00–7.20 a year to store on board the tankers. To purchase the
crude oil, it would cost USD 94 m (2 million barrels * USD 47.00/barrel) while the revenues received from selling the oil forward would be USD 120 m (2 million barrels *
USD 60.00/barrel). This would result in a gross profit (before expenses) of USD 26 m.
This period coincided with a substantial collapse in freight prices and interest rates,
which made the transaction attractive. The cost of using the vessel, as a storage facility
would be USD 14.4 m (2 million barrels * USD 7.20), while the cost of financing the
purchase of the oil and the hire of the ship would be USD 271,000 (USD 94 m + USD
14.4m * 0.25%)3 . The net profit would therefore be USD 11,329,000 (USD 26 m − USD
14.4 m − USD 271,000). It may be possible that the profits were even higher if BP had
extracted the oil from their own facilities at a marginal cost of production lower than the
spot price. This is clearly an attractive profit and so should encourage other participants
to execute a similar trade. This would force the spot price up and the forward price
down, removing the profits. By April 2009, the amount stored offshore had risen to
100–120 m barrels on 56 ships; in a normal year about 5–7 tankers would be used in this
manner. However, the number of barrels stored in this manner had fallen to 50–60m by
September 2009 as the differential between spot and 12-month forward prices had fallen
to USD 5.45 making the trade less attractive.
Technical or quality arbitrage
Trafigura (2015) gives an example of this. In the USA, gasoline typically contains some
amount of ethanol. It may be possible for an entity to source and blend the individual
components to earn an attractive margin.
3
An actual / actual day basis is assumed.
32
2.5
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
COMMODITY SWAP VALUATION
To illustrate the mechanics of a commodity swap, consider the following crude oil
example.
Maturity
Frequency of cash flow exchange
Notional amount
Floating price
Fixed price
One year
Monthly
10,000 barrels
Near-dated WTI futures price on settlement date
USD 57.59
Irrespective of the underlying asset class, the price of any swap is the fixed component of the transaction.
To calculate the theoretical price of a crude oil swap, the starting point is to appreciate that all swaps should be considered an equitable exchange of cash flows on the
day they are traded. That is, the present value of the expected payments must equal the
present value of the expected receipts. A reader new to the world of swaps may find this
odd as it suggests the transaction does not have any profit. However, note that the fair
price of the swap was described in terms of expected cash flows. Profits and losses will
arise as actual payments are crystallised – these could be substantially different than
those expected at the start of the transaction.
Since we are trying to solve for an unknown fixed rate, our analysis of swap prices
starts with the floating or variable cash flows. The aim on the floating side is to calculate
the present value of the future cash flows, which are linked to a series of yet to be determined prices. To derive an initial value for each of the floating cash flows it is market
practice to use the futures curve that prevails on the transaction date.
Table 2.1 illustrates all the cash flows associated with the swap.
The starting point is the fourth column that lists the applicable futures prices (floating prices) for the underlying commodity. Although these values are based on the current futures curve, an element of interpolation is needed, as the futures settlement date
will differ from that of the swap. These floating prices are then applied to the notional
amount (i.e. 10,000 barrels) to determine the magnitude of the floating cash flows (fifth
column). Although Table 2.1 approaches this in a different way, the next step would be to
present value and then sum the floating cash flows. The fixed price could then be solved
for iteratively. It is the single value, which when applied to the notional amount will
return a set of present-valued cash flows whose sum equals that of the present-valued
floating cash flows. The result will be a structure where the initial net present value
(NPV) of the cash flows is zero (column seven).
By looking at the single fixed rate in relation to the floating prices, the former is
a weighted average of the latter. Viewing the fixed rate in this manner gives an insight
into the essence of swaps. Ignoring any underlying economic exposure, an entity paying
fixed must believe that over the life of the swap they will receive more from the floating
cash flows. In other words, the fixed rate payer must believe that actual crude futures
prices will rise faster than those currently implied by the market. The opposite would
33
Derivative Valuation
TABLE 2.1 Swap cash flows on trade date.
Payment
dates
Fixed
price
Fixed
cash flows
Floating
prices
Floating
cash flows
Net
cash flows
Discounted
cash flows
20 Jan
21 Feb
20 Mar
20 Apr
22 May
20 Jun
20 July
21 Aug
20 Sept
20 Oct
20 Nov
20 Dec
57.59
57.59
57.59
57.59
57.59
57.59
57.59
57.59
57.59
57.59
57.59
57.59
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
575,884.80
56.20
56.80
57.22
57.62
57.87
57.96
57.97
57.96
57.92
57.88
57.86
57.82
−561,966.67
−568,025.00
−572,238.71
−576,228.57
−578,654.55
−579,600.00
−579,735.48
−579,564.52
−579,224.14
−578,837.50
−578,566.67
−578,214.29
13,918.14
7,859.80
3,646.09
−343.77
−2,769.74
−3,715.20
−3,850.68
−3,679.71
−3,339.34
−2,952.70
−2,681.86
−2,329.48
13,909.21
7,843.55
3,633.92
−342.12
−2,752.76
−3,687.57
−3,816.69
−3,641.67
−3,299.93
−2,913.25
−2,641.71
−2,290.97
be true of a fixed rate receiver; that is, they expect actual futures prices to be below those
currently implied by the market.
In this example, the cash flows are present valued using interest rates derived from
the interest rate swap market. The present valuing is done in terms of discount factors,
which have a value between 0 and 1. The discount factor can be applied to a future cash
flow using the following simple relationship:
Present Value = Future Value x Discount Factor
In its simplest form a discount factor can be represented by the following formula:
Discount factor =
1
(1 + zero coupon interest rate)n
Where n is the number of years from the effective date to the cash flow settlement date
Where the period relates to less than 1 year the formula becomes:
Discount factor =
1
1 + (zero coupon interest rate x
days
)
basis
Where ‘days’ is the number of actual days in the cash flow period and ‘basis’ is
either 360 or 365 depending on the money market convention associated with the swap’s
currency4 .
4
USD, EUR will take a 360-day convention, GBP will be 365.
34
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The interest rate used to calculate a discount factor should be a zero coupon in style
and should have the same degree of credit risk as the cash flow to which they will be
applied. A zero coupon instrument pays no cash flow until maturity with the buyer’s
return usually in the form of a capital gain. However, interest-bearing deposits may also
be in zero coupon form if they only have two cash flows – the initial and final movement
of funds. This means that an investor’s exact return for a given holding period can be
calculated. This zero coupon return is different than that offered by an interest-bearing
instrument that pays a series of cash flows prior to maturity. Although a yield to maturity can be calculated for an interest-bearing instrument, it would not be a true measure
of the overall return as this measure requires the interim interest payments to be reinvested at the yield that prevailed at the start of the transaction. This is referred to as
reinvestment risk.
One problem with zero coupon yields is the lack of available market observations
and as such an analyst is often forced to use mathematics to transform the yield on an
interest-bearing instrument into a zero coupon equivalent. A comprehensive treatment
of this subject can be found in Galitz (2013).
2.5.1
Single and dual curve discounting
Broadly speaking there are two approaches that could be used to value a swap: LIBOR
or Overnight Index Swap (OIS) discounting. To illustrate the difference between the
two techniques, consider the following stylised example. Examples of ‘real world’ swap
transactions will be illustrated in subsequent chapters.
Notional amount:
Maturity:
Fixed price payer:
Floating price payer:
Fixed price:
Floating price:
Frequency of payments:
10,000 barrels
Six months
Client
Bank
USD 51.44
Average of front month crude oil futures
settlement price
Monthly for both fixed and floating
At the time of writing the ICE Brent futures contract would cease trading on the
last business day of the second month that precedes the relevant contract month. For
example, the March contract expires on the last business day of January. So, based on
Table 2.1, the first cash flow on 20 January will reference the average of the maturity that
was the front month contract for that period. From the trade date until the 29 December
this would have been the February maturity; the period from then until the cash flow
settlement date would reference the March contract.
For the purposes of the following examples, suppose the following interest rates are
observed in the market. For ease of illustration each month is assumed to have 31 days.
35
Derivative Valuation
TABLE 2.2 LIBOR rates and their associated discount factors
Months
LIBOR rate
Discount factor
1
2
3
4
5
6
1.00%
1.25%
1.50%
1.75%
2.00%
2.25%
0.999140
0.998925
0.998710
0.998495
0.998281
0.998066
The applicable crude oil futures prices appear in Table 2.3.
TABLE 2.3 Crude oil futures prices
2.5.1.1
Month
Price (USD/barrel)
1
2
3
4
5
6
50.00
50.25
50.50
50.75
51.00
51.25
LIBOR discounting
It was argued earlier that on the trade date a swap should be an equitable exchange
of cash flows. Expressed more formally this means that the present value (PV) of the
cash flows to be paid must equal the present value of the cash flows to be received. If
the swap were structured so that the fixed cash flows are received and the floating cash
flows paid, this would suggest that:
PV of fixed cash flows less PV of floating cash flows = 0
The PV of the fixed cash flows (FXD) can be calculated using a series of discount
factors (DFn ):
=FXD ∗ DF1 + FXD ∗ DF2 + FXD ∗ DF3 + FXD ∗ DF4 + FXD ∗ DF5 + FXD ∗ DF6
=FXD ∗ (DF1 + DF2 + DF3 + DF4 + DF5 + DF6 )
(2.5)
The PV of the floating cash flows (FLTn ) can be calculated in a similar manner:
= FLT1 ∗ DF1 + FLT2 ∗ DF2 + FLT3 ∗ DF4 + FLT5 ∗ DF5 + FLT6 ∗ DF6
(2.6)
36
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
It may be obvious to state that these equations can only be solved if we know the
value of the various inputs. Arguably, the easier input to determine is the value of the
discount factor, which can be derived from observed LIBOR rates.
The next step is to calculate the floating cash flows. However, the cash flows that
will be settled will be based on prices that prevail at some future date. The challenge is
to determine a series of values that will apply at the inception of the transaction. Rather
than forecast what these cash flows are likely to be the swap trader could hedge these
cash flows using futures. In our example the floating cash flows are to be paid by the
trader and so they could be hedged by buying a series of futures contracts. So, if the price
of crude oil rises and the trader is faced with increased payments on the floating leg of
the swap, an offsetting long position in a futures position should show a corresponding
profit. On this basis the convention is to use prevailing futures prices to assign a value to
the floating swap cash flows on the trade date. This highlights a key aspect of derivative
valuation – the price of a transaction is driven by the cost of hedging the underlying
exposure.
Since the swap has to have a zero market value on the trade date the only remaining
unknown factor is the fixed price. With a little bit of rearrangement of equations (2.5)
and (2.6) it is possible to solve for a fixed rate (FXD) that achieves this objective. The
single fixed price that will apply to the swap is calculated as USD 51.44 expressed to two
decimal places.
To illustrate this concept, the following table shows the value of the crude oil swap
as of the trade date.
The fixed cash flows are calculated as:
Number of barrels x fixed price
The example in Table 2.4 was calculated using Excel and so the fixed price on a per
barrel basis is expressed to six decimal places: USD 50.624843.
The first fixed cash flow in column 2 is calculated as:
10,000 barrels × USD 50.624843 = USD 506,248.43
TABLE 2.4 The valuation of a commodity swap under LIBOR (‘single curve’) discounting.
Months
1
2
3
4
5
6
Fixed
cash flows
PV of fixed
cash flows
Floating
cash flows
PV of floating
cash flows
506,248.43
506,248.43
506,248.43
506,248.43
506,248.43
506,248.43
PV TOTALS
505,812.87
505,704.10
505,595.37
505,486.69
505,378.06
505,269.47
3,033,246.57
−500,000
−502,500
−505,000
−507,500
−510,000
−512,500
−499,569.81
−501,959.70
−504,348.55
−506,736.38
−509,123.18
−511,508.95
−3,033,246.57
NET PRESENT VALUE
0
37
Derivative Valuation
The present value of the fixed cash flows (column 3) is calculated as:
Fixed cash flow x discount factor
The first cash flow is calculated as:
USD 506,248.43 × 0.999140 = USD 505,812.87
The floating cash flows in column 4 are calculated as:
Number of barrels x futures price
The first floating cash flow is:
10,000 × USD 50.00 = −USD 500,000
The PV of the floating cash flows (column 5) is calculated as:
Floating cash flow x discount factor
The first calculation is:
USD 50,000 × 0.999140 = −USD 499,569.81
Over time the market factors used to establish the initial value of the swap will
evolve. As a result, it is perhaps inevitable that the swap will have value for one of the
market participants at the expense of the other. To establish how this value evolves, it
is possible to use the principles just outlined to determine the value of the swap. To
illustrate this concept, suppose that crude oil prices rise instantaneously by USD 1.00
across all the applicable maturities (all other things being equal). This will increase the
magnitude of the floating cash flows to be paid resulting in a new net present value of
−USD 59,916.17.
The swap also has exposure to the passage of time. Assuming all else being equal
then after one month the swap will only have five pairs of cash flows remaining. The
PV of the fixed receipts will be USD 2,527,433.69 while the PV of the floating payments
will be USD 2,533,675.75 resulting in a net present value of −USD 6,243 to the bank.
The structure also has an exposure to the change in interest rates but in this relatively
short-dated example the impact is negligible.
2.5.1.2
Overnight index swap discounting
Derivative credit risk exists for contracts that are ‘in the money’ (i.e. profitable). Given
that derivatives are bilateral contracts, the credit risk could switch between the contracting entities depending on how the profitability of the contract evolves. Counterparty
default does not necessarily lead to a loss. It would only do so if the contract were valuable to the non-defaulting entity.
To mitigate the counterparty credit risk associated with a derivative transaction it
is common for the contracting parties to agree to exchange collateral. Collateral agreements (e.g. the ISDA credit support annex) typically stipulate that interest will be paid
on monies held as collateral. The most common interest rate that is used in this respect is
38
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
one that references an overnight rate. For USD contracts this could be the Fed Funds rate
while for EUR contracts it would be the European Overnight Index Average (EONIA).
Since these interest rates are very short term, the market perceives them to have a very
low degree of credit risk.
This has led to the use of overnight index swap (OIS) rates rather than LIBOR
for valuation purposes. An OIS swap is an exchange of fixed cash flows for a series
of floating cash flows where the latter reference an agreed overnight index. Since an
OIS rate is perceived to have a lower credit risk, they should trade below the equivalent
LIBOR.
To give a sense of why an OIS would work consider a situation where a market
participant has a crude oil swap with a residual maturity of one year and a notional of
10,000 barrels. For ease of illustration we will assume that the fixed price was originally
set at USD 50.00/bbl. and that the floating cash flows are paid monthly referencing a
futures contract. In this example the fixed price is received annually so there is only
one of these cash flows remaining. The trader decides to offset this position with a
swap that has the same characteristics but is traded at prevailing rates. If the fixed
price is now USD 45.00/bbl., then the trader has been able to lock in a profit of USD
5.00/bbl., or a total of USD 50,000. The floating cash flows will cancel out and the
profit is the difference between the two fixed prices. If 12-month LIBOR rates are,
say, 5% then the mark to market of this transaction is USD 47,619 (USD 50,000/1.05).
To protect the value of this position the trader takes collateral on the original ‘receive
fixed’ position equal to this amount and agrees to pay the 12-month OIS rate. If this
rate is, say, 4% over the course of the year, then the collateral will grow to reach USD
49,523 (USD 47,619 x 1.04). At the maturity of the transaction the trader expects
to receive USD 50,000 on a net basis but taking collateral discounted at LIBOR but
accruing at an OIS rate will not generate sufficient proceeds if the counterparty were
to fail. The correct amount of collateral that the trade should demand is USD 48,077
(USD 50,000/1.04); paying the OIS rate on this amount of collateral will generate USD
50,000 at maturity.
This suggests that the correct rate at which cash flows at which a collateralised deal
should be discounted is the OIS rate. Referring to the discounting example in Table 2.5
the discount factors are now calculated using OIS rates rather than LIBOR. The applicable formula to calculate the discount factor is the same as LIBOR except that an OIS
rate is used.
TABLE 2.5 Overnight Index Swap rates and their
associated discount factors.
Months
OIS rate
Discount factor
1
2
3
4
5
6
0.50%
0.75%
1.00%
1.25%
1.50%
1.75%
0.999570
0.999355
0.999140
0.998925
0.998710
0.998495
39
Derivative Valuation
The cash flows on the swap on the trade date appear in Table 2.6.
TABLE 2.6 The valuation of a commodity swap under LIBOR (‘single curve’) discounting.
Months
1
2
3
4
5
6
Fixed
cash flows
PV of fixed
cash flows
Floating
cash flows
PV of floating
cash flows
506,248.43
506,248.43
506,248.43
506,248.43
506,248.43
506,248.43
PV TOTALS
506,030.56
505,921.69
505,812.87
505,704.10
505,595.37
505,486.69
3,034,551.28
−500,000
−502,500
−505,000
−507,500
−510,000
−512,500
−499,784.81
−502,175.68
−504,565.51
−506,954.32
−509,342.10
−511,728.85
−3,034,551.28
NET PRESENT VALUE
0
Source: Author’s calculations
Note that on the trade date the net present value of the swap is still zero and that
the fair price of the swap is USD 50.62, although the present value of the individual
cash flows is different. This will manifest it most clearly when futures prices change.
If futures prices were to increase by USD 1.00/bbl. across the entire curve, then under
LIBOR discounting the swap will lose USD 59,916.17. Based on the same scenario, the
OIS discounting would show a loss of USD 59,941.94. Although it may seem a relatively
small amount, over a large portfolio of swaps that perhaps have larger notional amounts
and longer maturities, this impact could be pronounced.
2.6
PRINCIPLES OF OPTION VALUATION
Probably one of the most documented areas of finance is that of option pricing. Since
the aim of this chapter is to give readers a basic understanding of where the value of a
derivative instrument comes from, the analysis will avoid excessive discussions on the
mathematics of options and concentrate on the intuition. Those interested in understanding the mathematics are recommended to refer to a variety of texts such as Galitz
(2013), Tompkins (1994) and Natenberg (2014). A more recent and commodity specific
text has been written by Geman (2005)
2.6.1
Black Scholes and Merton
An option’s premium is primarily a function of:
▪ The expected payout at maturity
▪ The probability of the payout being made
Although at an intuitive level these concepts are easy to understand the mathematics behind the principles is often complex and off-putting to many readers. To determine
the premium on an option, a variety of inputs are required as well as is an appropriate
model. Under a Black Scholes and Merton (BSM) framework, those inputs will include
▪ The spot price,
▪ The strike price,
40
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Time to maturity,
▪ The cost of carrying the underlying asset as a hedge (i.e. any income earned through
holding the underlying asset less any expense incurred),
▪ The implied volatility of the underlying asset.
The following input values will be used for the subsequent examples.
Underlying commodity
Gold
Spot
Strike
Time to maturity
Six-month LIBOR
Six-month lease rate
Six-month forward price
Implied volatility
USD 1,400
USD 1,400
Six months
3%
0.10%
USD 1,420.44
15%
We will assume that the option is European in style and given the input values noted
above, a Black Scholes Merton model returns a premium of USD 69.38 per troy ounce
(option premiums are quoted in the same format as the underlying asset).
Note that at first glance this option appears to have been struck at-the-money, as
the spot price and the strike are the same. However, since this option is European the
model will value the option relative to an equivalent underlying price. Since the option
cannot be exercised for six months, the equivalent underlying is not the spot price but
the six-month forward. So, the option is in-the-money; the call buyer has the right to
buy at USD 1,400 when the underlying market is USD 1,420.44.
An option premium can be decomposed into two elements: time value and intrinsic
value. The intrinsic value can be thought of as the amount of profit the buyer would
make if they were to exercise the option straightaway. However, the term is defined in
such a way that it does not consider the initial premium paid. The intrinsic value for a
call option can be expressed in the following manner.
Intrinsic value of a call = MAX (0, Underlying price − strike)
For put options it is expressed as
Intrinsic value of a put = MAX (0, Strike − Underlying price)
In both expressions the underlying price is the forward price for European style
options and spot for American style. In the original definition of intrinsic value prior to
maturity for a European option, the difference between the underlying forward price
and the strike would need to be present valued since exercise of the European style
option could only take place at expiry and the premium is quoted as a present value.
However, when analysing options intuitively, traders would probably disregard this
aspect.
Derivative Valuation
41
Time value is the extra amount that the seller charges the buyer to cover them for
the probability of future exercise – a sort of uncertainty charge. Time value is not paid
or received at expiry of the option as for buyers it will fall to zero over the option’s life.
When an option is exercised the holder will only receive the intrinsic value. Time value
is only paid or received upon the sale or purchase of an option prior to maturity. This
explains why early exercise of an option is rarely optimal. If a purchased option is no
longer required it would usually be more efficient to neutralise the exposure by selling
an opposite position in the market. In this way they would receive both intrinsic and
time value. If the option were exercised, they would simply receive the intrinsic value.
An understanding of the two option premium components is vital to understand
the logic of interbank transactions. Intrinsic value is principally driven by movements
in the underlying price, while time value is primarily a function of time to expiry and
implied volatility.5
This means that when trading options the three main factors that need to be considered are:
▪ Movements in the underlying price.
▪ The passage of time.
▪ Movements in implied volatility.
2.6.2
The Black model
One of the ways to approach the valuation of commodity options that reference an
underlying future is to use the Black model (1976) for options on futures. This model is
used extensively in the interest rate options market for valuing caps, floors, and swaptions. In a Black Scholes Merton model an analyst would typically input the spot price of
the asset into the model, which combined with the ‘carry’ parameters (e.g. interest rates
and any income the underlying asset would generate) would determine the implied forward price. Based on the implied forward, the model would then calculate the option’s
fair value. One of the main attractions of the Black model is its simplicity as the spot and
net carry parameters are replaced with the observed forward or futures price. This would
mean that the parameters required to price an option on a crude oil future would be:
▪ Price of the future.
▪ Strike.
▪ Option’s time to maturity.
▪ The implied volatility of the future’s price.
2.6.3
Bachelier model
One of the features of the Black Scholes Merton (BSM) and Black models is they assume
that the underlying prices are lognormally distributed. From a price perspective, this
means the models do not allow for the possibility that prices could turn negative.
5
See Tompkins (1994).
42
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
For conventional financial assets such as shares, bonds, and foreign exchange this is
a reasonable assumption; when a company goes bust, their shares will stop trading
rather than turn negative.
As ever, commodities are different. Negative prices have been experienced in power
and gas (see chapters 7 and 8) and calendar spread trades can flip between positive and
negative values. The issue of negative prices and index values came more to the fore
in 2020 as a result of activity in the shipping market (e.g. Baltic Capesize Index)
and in April 2020 for the first time, crude oil (see see chapter 6). This brought into
question the use of models such as Black 1976 for valuing and risk managing options
on futures.
An alternative to this model is the Bachelier model, which dates from around 1900.
Some of the key characteristics of the Bachelier model are:
▪ It assumes that the underlying price follows a normal rather than lognormal distribution and so allows for the possibility of negative prices.
▪ Volatility is expressed as a monetary value rather than as a proportional percentage
value. For example, assume an asset is trading at a price of USD 10.00 with an
implied volatility of 10% per day. This would suggest that the asset is expected
to trade with an approximate range of prices between USD 9.00–11.00. If the
price drops to USD 5.00, but the implied volatility remains the same, then the
approximate range of prices would be +/− 50 cents. Using the same set of prices,
a Bachelier model would express volatility as a dollar value, say, USD 1.00. So, at
USD 10.00 the range of prices would be similar. But if prices fell to USD 5.00 then
under a Bachelier framework, the range of prices remains at +/− USD 1.00. At very
low prices (e.g. USD 1.00), but with volatility expressed as a dollar value (e.g. USD
20.00), it means that the model implies a range of values from −USD 19.00 to
+USD 21.00.
▪ Since the model allows for negative prices, OTM puts will be valued relatively
higher using a Bachelier model rather than a BSM framework. Since the premium
reflects the expected payout the Bachelier model reflects, this pushes the price
relatively higher than the BSM framework. For example, using a simple Bachelier
model a three-month put option with a zero strike, an underlying price of zero,
and a dollar volatility of USD 40.00 returns a premium of USD 7.98. A Black 1976
model returns a premium value of USD 0.00.
▪ A BSM framework will value OTM calls relatively higher than Bachelier mainly
due to the differences in shape between a lognormal and normal distribution.
▪ The option’s delta will not be the same as under a BSM framework and so the risk
management of the option position will be different. The ATM put option valued
in the last point had a delta of exactly 0.5, which is less likely than under a BSM
framework.
▪ Risk (2020a) also argues that Bachelier tends to price shorter-dated options more
effectively than longer-dated options as they can capture a sudden downward move
in price.
▪ Risk (2020a) also points to the volatility skew associated with options valued under
a Bachelier framework arguing that the flatter the skew the better the model is at
capturing the behaviour of the market.
43
Derivative Valuation
2.6.4
Put-call parity: the theory
Although pricing models provide one linkage between the underlying price, the forward
rate and the option premium, the concept of put-call parity is an alternative representation. Tompkins (1994) provides a detailed analysis of the concept.
Put-call parity is a concept that attempts to link options with their underlying assets
such that arbitrage opportunities could be identified. The conditions of put-call parity
will hold if the strike, maturity, and amount are the same.
Although put-call parity varies according to the underlying asset, in its simplest
form it can be represented by the following expression:
C−P=F−E
Where
C = Price of a call option
P = Price of a put option
F = Forward price of the underlying asset
E = Strike rate for the option
Strictly speaking the right hand side of the expression would need to be present
valued as the strike and the forward price relate to a future time period, whereas the
call and put premiums are expressed in present value terms.
Although put-call parity was designed to identify price discrepancies between markets, it is possible to adapt the formula so it can be used from a strategy perspective. In
this case the expression can be written as:
C−P=F
Where F is redefined as a position in the underlying, in this case a forward position.
Each of the symbols in the expression can be annotated with either a ‘+’ or a ‘−’ to
indicate either a buying or selling position, respectively. So:
+C − P = +F
That is, buying a call and selling a put is equivalent to a long position in the underlying. Although this may seem a very dry concept, the relationship is frequently used in
financial engineering to create innovative structures.
2.6.5
Put-call parity: the application
One way in which put-call parity could be applied is in the construction of hedges. Let
us take an example of a client looking to buy a commodity forward (e.g. an automotive
hedger). If the hedger was looking to buy the commodity at a strike rate equal to the current forward rate, then the price of a call and a put will be the same. Not unreasonably,
he may wish to achieve a strike rate E that is less than the current forward rate F, which
44
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
would suggest that for the relationship to hold the price of the call will increase relative
to the put. The challenge therefore becomes to obtain a more favourable hedging rate
by cheapening the value of the call option.
This can be achieved using three possible strategies:
▪ Buy a call on a notional amount of N and sell a put with a notional amount of 2N
(sometimes referred to as a ratio forward).
▪ Buy a reverse knock out call option with the barrier set at a level that is unlikely to
trade.
▪ Buy a short-dated call option and finance this by selling a longer-dated put option
at the same strike.
2.7
MEASURES OF OPTION RISK MANAGEMENT
An option’s ‘Greeks’ are measures of market risk that have been developed over time to
help market participants manage the risk associated with the instruments. Each of the
option model inputs has a related Greek whose value can be calculated numerically (see
Tompkins, 1994 or Haug, 2007) or derived by perturbing the appropriate option model.
Most of the Greeks look at how a change in the relevant market input effects the value
of the option premium, that is, they are first order effects; measures such as gamma are
second order in nature.
2.7.1
Delta
The delta of an option looks at how a change in the underlying price will affect the
option’s premium. In one sense, delta can be thought of as the trader’s directional exposure. There are a variety of different definitions of delta:
▪ The rate of change of the premium with respect to the underlying.
▪ The slope of the price line.
▪ The probability of exercise.
▪ The hedge ratio.
The most traditional way of defining delta is the rate of change of the premium with
respect to the underlying. Using this method, with a known delta value, the analyst can
see how a small change in the underlying price will cause the premium to change. In
formula terms it is expressed as:
Delta =
Δ Premium
Δ Underlying price
So, if an option has a delta value of 0.575, we can say that the premium should
change by 57.5% of the change in the underlying. Using the gold option analysed in
Section 2.6.1, if the underlying price rises by USD 0.50 from USD 1,400 to USD 1,400.50,
the premium should rise by USD 0.2875 cents from USD 69.38 to USD 69.6675. Using
45
Derivative Valuation
TABLE 2.7 Premium against underlying price for a call option. Associated
delta values shown in column 3.
Underlying price (USD)
1,325
1,350
1,375
1,400
1,425
1,450
1,475
1,500
1,525
Premium (USD)
Delta value
33.89
44.00
55.84
69.38
84.56
101.29
119.43
138.83
159.33
0.37
0.44
0.51
0.58
0.64
0.70
0.75
0.80
0.84
an option spreadsheet, the premium moves to USD 69.6638. One of the reasons why the
prediction is not accurate is that delta only applies to small changes in the underlying
price.
Delta will be either positive or negative depending on whether the option has been
bought or sold or is a call or a put. For a purchased call or sold put the associated delta
value will be positive. This is because in both instances the value of the option will rise
if the underlying price increases – a positive relationship. For a purchased put and a
sold call the delta value for these options will be negative. This is because a rise in the
underlying price will cause the options to fall in value – a negative relationship. Both
statements can be verified by looking at the payoff diagrams shown in Figure 1.2.
The second way of illustrating the concept of delta is to see how the value of the
premium changes as the underlying spot price changes (all other things being equal).
Table 2.7 shows the premium and delta values for a range of underlying prices for our
example call option.
The resultant graph (Figure 2.3) from plotting the values in the first two columns
is shown below.
180
160
140
120
100
80
60
40
20
0
1,325
1,350
1,375
1,400
1,425
1,450
1,475
FIGURE 2.3 Premium (Y axis; USD/oz.) against price (prior to maturity).
1,500
1,525
46
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The second definition of delta is the slope of the price line prior to maturity (e.g.
premium against price as per Figure 2.3). Delta is the slope of a tangent drawn at
any point on the curve. This also suggests that delta should only be used to measure
small changes in the underlying price. As the premium-price relationship is convex in
nature using delta to determine the impact of significant price movements would give
an incorrect estimate.
Although not strictly true, delta is often interpreted as the probability of exercise.
The application of the term in this definition is not uniformly accepted but it certainly
provides a useful rule of thumb.
Delta can be measured in a variety of ways:
▪ As a number whose value will range from zero to one
▪ As a percentage
▪ As a delta equivalent figure
The most common way of expressing delta is a value ranging between zero and one,
with a positive or negative sign associated. Out-of-the-money options will have a delta
that ends towards zero while in-the-money option deltas tend towards one. It would
be incorrect to say that all at-the-money options have a delta of 0.50 but the value will
certainly be close. A plot of the delta values shown in Table 2.7 against the spot price of
gold is shown in Figure 2.4.
An alternative method is to express it as a percentage number, which was also illustrated earlier. The third way is to express it as a delta equivalent value. This technique
compares the market risk position of the option with that of an equivalent underlying
asset. So if we were to buy a call option on a notional amount of 50,000 ounces with a
delta value of 0.58, the option has the same degree of market risk (for small changes in
the underlying asset) as 29,000 ounces (50,000 × 0.58)
The concept of delta can be extended to option positions. If we were to add another
bought call option to the above position with a notional of 60,000 ounces and a delta
of 0.40, this individual position would have a delta equivalent value of 24,000. On a
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1,325
1,350
1,375
1,400
1,425
FIGURE 2.4 Delta (Y axis) versus spot price of gold.
1,450
1,475
1,500
1,525
47
Derivative Valuation
net basis the entire option position has a delta equivalent exposure of 53,000 ounces
(24,000 + 29,000). The option trader’s position would have the same exposure to small
movements in the underlying price as long physical holding of 53,000 ounces.
If the option trader chooses to neutralise this exposure, he could take an offsetting
position in other options or the underlying instrument. He may wish to do this if he
is seeking to benefit from a change in another market factor such as implied volatility.
In the previous example the trader is delta positive so his option position will rise in
value if the underlying market price rises and vice versa. To neutralise the exposure he
could therefore sell 53,000 ounces of gold in the underlying market or trade options in
different combinations to achieve a net delta exposure with which he is comfortable.
If the net delta position is reduced to zero, this is described as delta neutrality. This
technique is referred to as delta hedging and demonstrates the use of delta as hedge
ratio.
If the delta neutralising trades are executed in the underlying market the trader
is faced with making a choice between trading in either the spot or forward market.
If the option position is European, then the forward market should be chosen. This is
because the European option can only be exercised upon maturity and so the appropriate equivalent market is the forward market. However, some traders may choose to
hedge in the spot market, and then hedge the resultant interest rate risk (recall that
interest rates are one of the factors that determine the difference between the spot and
forward price).
We have seen that delta changes as a function of the underlying price, but will also
evolve over time. For example, as the option approaches maturity the price line shown
in Figure 2.3 above will become more linear in nature and at maturity will resemble the
‘hockey stick’ shape that was illustrated in Figure 1.2. As a result, all other things being
equal, the delta of the option will tend towards zero or one. This evolution of delta with
respect to time is referred to as ‘delta bleed’ and represents another risk to a trader. This
is illustrated in Figure 2.5.
1
0.9
Delta tends
to one
0.8
0.7
0.6
0.5
0.4
Delta tends
to zero
0.3
0.2
0.1
0
1,325
1,350
1,375
1,400
FIGURE 2.5 ‘Delta bleed’ of a call option.
1,425
1,450
1,475
1,500
1,525
48
2.7.2
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Gamma
Gamma is defined as the rate of change of delta with respect to the price and is therefore
a second order function.
Δ Delta
Gamma =
Δ Premium
However, this definition is not helpful in trying to understand the practical aspects
of gamma. Over the course of his career the author has collected a few alternative definitions:
▪ The speed with which a delta hedged position becomes unhedged.
▪ The exposure to actual volatility in the market.
▪ The rate of change of a trader’s profit or loss.
▪ The exposure to significant changes in the underlying price.
Gamma is positive for option buyers (irrespective of whether the option is a call or
a put) and negative for sellers. Using the gold call option analysed in section 2.6.1, for
an underlying price of USD 1,400, the option pricing model returned a value for delta of
0.58 and 0.0026 for gamma. If we take an exposure of 50,000 ounces the delta exposure
is equivalent to owning 29,000 ounces of the underlying of the asset. We will assume the
trader hedges this exposure by selling an equivalent amount in the underlying market.
However, although the position is initially delta neutral, a movement in the underlying will lead to a change in delta and the position will no longer be 100% hedged. The
gamma measure can be used to predict the magnitude of any move in delta for a small
move in the underlying price. So in this instance we can say that for a small increase in
the underlying market price, the delta will move to 0.5826 (e.g. 0.58 + 0.0026) and for a
small move down it would move to 0.5774 (e.g. 0.58 − 0.0026).
2.7.3
Theta
Theta reflects the impact of the passage of time on the value of the option and is
defined as:
Δ Premium
Theta =
Δ Time
Theta is positive for sellers and negative for buyers. This reflects the fact that for
buyers, options will experience time decay. In other words, the buyer will pay a premium
and will acquire a contract that has time value and possibly intrinsic value. However,
over time, the time value element will fall to zero. Using the original call option example,
where the spot and strike are USD 1,400 and holding all the other parameters constant,
the effect of the passage of time on the option is reflected in Table 2.8.
Table 2.8 shows that for the buyer of an option its value will decrease as time passes
and that the rate of decay will accelerate as the option approaches maturity. The opposite
will hold true for sellers of options. Theta can be thought of as the slope of the price line
with respect to the passage of time (column 2 in Table 2.8). Theta will be small initially
and will increase with the passage of time (column 3 of Table 2.8). Note that in Table 2.8
the change in the value of the option with respect to time is measured in units of 0.1 of
49
Derivative Valuation
TABLE 2.8 The effect of the passage of time on the value of an option.
Time to expiry (years)
Premium (USD per troy ounce)
Change in premium
103.94
97.65
91.10
84.26
77.05
69.38
61.10
51.98
41.53
28.53
8.58
−
6.29
6.55
6.84
7.21
7.67
8.28
9.13
10.45
13.00
19.95
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.01
a year (about 37 days). It is more common to express theta in terms of the impact of a
single day.
2.7.4
Vega
Vega is defined as the change in the option premium for a 1% change in an options
implied volatility. It is expressed as:
Vega =
Δ Premium
Δ Implied volatility
Although most people are happy with the concept of volatility in general terms,
some magnitude of change in market prices with no direction suggested, the notion of
implied volatility sometimes remains difficult to grasp.
Consider first the related concept of historical volatility, which as the term suggests, measures the magnitude of movement of the underlying asset over a historical
period, measured in per cent per annum. The statisticians would describe this as a standard deviation. However, as is often cited in financial markets ‘that past performance
is not a guide to future performance’, and there is a need for a more forward-looking
measure of volatility.
One of the most popular ways of trying to explain the concept is to describe as the
volatility implied in an observed option price. The rationale for this is that if one looks
at all the option pricing model inputs the only real unknown is the implied volatility.
Terms such as the spot rate or strike are either easily observable or negotiated as part
of the option contract. Since it is the only unknown factor, the traditional definition
suggested taking an observed option premium, inserting the value into the option pricing model, ‘running the model in reverse’ and backing out the volatility implied by this
price. However, this argument traditionally ceases at this point and does not address the
obvious circular argument; where did the market maker quoting the observed option
price obtain his volatility input? It also gives no real feel for what the measure is trying to achieve. Tompkins (1994, p. 139) describes it as ‘the risk perceived by the market
50
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
today for the period up until the expiration of a particular option series’. However, since
no one can foretell the future with complete certainty, implied volatility is nothing more
than a best guess as to the future expected volatility of the underlying asset. As an anecdote, the author recalls teaching a class attended by an experienced option trader who
described the interbank options as the ‘market for guesses’. In the interbank market
option prices are quoted in terms of a bid-offer implied volatility number. Since implied
volatility is the only truly unknown variable this is the factor that it traded on an interbank basis. Trading strategies that aim to exploit movements in implied volatility are
covered in chapter 3.
The main confusion regarding implied volatility often surrounds the relationship
with movements in the underlying asset price. The delta value of an option is sensitive
to the magnitude of the implied volatility input and as the size of the volatility number increases the delta on the option will tend towards 0.50. This is because the range
of expected values that the underlying is expected to take at maturity is wider and so
intuitively, the ‘odds’ of the option being exercised tend towards 50/50. However, it does
not follow that as the underlying price starts to move that implied volatility must also
move as well. In an option model implied volatility and the spot price are two independent variables and as such, a movement in one factor can be independent of the other.
Although it may be reasonable to assert that current observed movements in the spot
price may alter a trader’s expectations on the distribution of spot values at maturity, the
point is that there is no inbuilt dependency. There is no ‘ruling’ that says if spot were to
change by X% that implied volatility must change by Y%.
The confusion over the relationship between spot and volatility is like that seen
between vega and gamma. Gamma was earlier introduced as a trader’s exposure to
actual volatility. If we saw a large value for gamma then the difference between two
delta values will be substantial for a given change in the underlying market. If the difference between two delta values is significant this means that the movement of the
spot prices was quite large, hence one could reasonably describe such a market as being
currently volatile.
Table 2.9 shows the effect of implied volatility on the premium for three variants of
the call option introduced in Section 2.6.1.
▪ An out-of-the-money option with a strike rate of USD 1,450.
▪ An at-the-money forward option with the strike rate is equal to the forward rate of
USD 1,420.
▪ An in-the-money option with a strike rate of USD 1,400.
As before all the other pricing parameters are held constant.
From Table 2.9 we can draw several conclusions:
▪ The relationship between an ATM option premium and implied volatility is proportional. Doubling the implied volatility will double the premium.
▪ If the option has no implied volatility an OTM and ATM option will have no premium, although the ITM option will have its intrinsic value. This also suggests that
volatility is arguably the defining factor that distinguishes an option from its underlying asset.
51
Derivative Valuation
TABLE 2.9 The impact of implied volatility on a variety of options with different degrees of
‘moneyness’.
Implied
Volatility
0%
5%
10%
15%
20%
25%
30%
OTM
option
premium
Change
in OTM
premium
ATM
option
premium
Change
in ATM
premium
ITM
option
premium
Change
in ITM
premium
0.00
8.67
27.00
46.36
65.99
85.70
105.43
8.67
18.33
19.36
19.63
19.71
19.73
0.00
19.73
39.46
59.18
78.88
98.55
118.20
19.73
19.73
19.72
19.70
19.67
19.65
20.00
31.29
50.07
69.38
88.80
108.25
127.69
11.29
18.78
19.31
19.42
19.45
19.44
▪ The relationship between an OTM option premium and the implied volatility is
also non-linear. Notice how a doubling of the implied volatility at low levels causes
the premium to rise by a much greater magnitude.
Option buyers are said to be long vega (i.e. they will benefit if implied volatility
rises), while sellers are short vega (i.e. they will suffer if implied volatility rises, but will
benefit if it falls). The logic behind this lies in the pricing formula for options. Recall
that the premium of an option is composed of the intrinsic and time value. The intrinsic
value of the option is driven by the difference between the price of the underlying and
the strike price. The main drivers of the time value are time and implied volatility. The
buyer of an option acquires both elements and all other things being equal, an increase
in implied volatility will increase the value of the option; a positive relationship.
2.7.5
Non-constant volatility
One of the assumptions of option valuation theory is the notion of constant volatility. This means that irrespective of the strike or tenor of the option the volatility input
should be of the same value. For example:
▪ A one-month option should be valued using the same volatility as, say, a 12-month
option.
▪ An OTM option will be valued using the same volatility as an ATM option.
Empirically, these conditions are not observed in practice. The two main exceptions
relate to the existence of skews and term structures.
2.7.5.1
The volatility skew
A volatility skew describes how different volatilities are used to value options of a given
maturity but with different strikes. A representative volatility skew for options on WTI
is shown in Figure 2.6.
52
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
38
Implied volatility (%)
36
34
32
30
28
26
24
22
P
25 UT
D
P
30 UT
D
P
35 UT
D
P
40 UT
D
PU
T
45
D
PU
T
AT
45
M
D
C
AL
40
L
D
C
A
35
LL
D
C
30 ALL
D
C
AL
25
L
D
C
20 ALL
D
C
15 AL
L
D
C
A
LL
10
D
C
AL
L
PU
T
20
D
15
D
10
D
PU
T
20
Relative strike
FIGURE 2.6 Volatilities with respect to strike price for 3-month options on WTI crude oil.
Source: Author, CME group
The ATM volatility is about 29% while low strike options (e.g. OTM puts with low
delta values, e.g. 10D put) are higher, and high strike options (e.g. OTM calls with low
delta values, e.g. 10D calls) are lower. This would be a case of the market being ‘skewed
to the downside’. One way of interpreting the shape of this curve is by the simple interaction of demand and supply. Recall the main inputs for an option valuation model
include the underlying price, strike, time to maturity, and implied volatility. All these
components apart from implied volatility are either known with reasonable certainty
or agreed with the counterparty. As a result, traders will express views on the component that is relatively uncertain – the implied volatility (IV). Suppose that the market
expects oil prices to fall; traders could profit from this view by buying OTM put options.
The option is traded with a low strike and so its premium will be relatively low. However, if there is an increase in demand its value expressed in volatility terms will increase
given the higher demand. Since this option will incur a premium it may well be that the
trader will finance this purchase by selling options in areas where he does not expect the
underlying to trade; in this case OTM calls with relatively higher strikes. To illustrate
the concept, consider the following market values.
▪ Three-month futures price = USD 55.73.
▪ ATM option valued with an IV of 29%. Premium =USD 3.1913.
▪ 10D call, strike rate of USD 66.59, valued with an IV of 26.47%. Premium = USD
0.3307.
▪ 10D put, strike rate of USD 45.06, valued with an IV of 35.80%. Premium = USD
0.5117.
For a given maturity the ATM option costs more in premium terms than either of
the two OTM options.
This skew relationship is not static and will evolve over time. It will also vary
between different commodity classes.
53
Derivative Valuation
Implied volatility (%)
19
18
17
16
15
14
P
25 UT
D
P
30 UT
D
P
35 UT
D
P
40 UT
D
PU
T
45
D
PU
T
AT
45
M
D
C
AL
40
L
D
C
35 ALL
D
C
30 ALL
D
C
AL
25
L
D
C
A
20
LL
D
C
15 AL
L
D
C
A
LL
10
D
C
AL
L
PU
T
20
D
10
D
15
D
PU
T
13
Relative strike
FIGURE 2.7 Volatilities with respect to strike price for 3-month options on gold.
Source: Author, CME group
20
D
15
D
10
D
P
25 UT
D
P
30 UT
D
P
35 UT
D
P
40 UT
D
P
45 UT
D
PU
T
AT
45
M
D
C
AL
40
L
D
C
35 AL
L
D
C
A
30
LL
D
C
25 ALL
D
C
20 ALL
D
C
15 AL
L
D
C
A
10
LL
D
C
AL
L
48
47
46
45
44
43
42
41
40
39
38
37
PU
T
PU
T
Implied volatility (%)
Figure 2.7 illustrates a typical skew shape for gold.
Many practitioners argue that gold trades more like a financial asset and often
incorporate gold desks within their foreign exchange trading activities. Here the relationship is more symmetrical and resembles a smile. One explanation of this shape is
that since it is traded as a quasi-currency, the market is split between a crash of the USD
and a crash of gold (currency code XAU). During the financial crisis of 2008, the implied
volatilities for high strike call options soared as gold moved towards USD 2,000/oz.
Figure 2.8 illustrates the relationship for US natural gas.
In this case there is a tendency for the market to be ‘skewed to the upside’ (i.e. OTM
calls tend to trade with relatively higher IVs than OTM puts). This may indicate that
the market expects the price of natural gas to rise over the coming months. A similar
Relative strike
FIGURE 2.8 Volatilities with respect to strike price for three-month options on US natural gas.
Source: Author, CME group
54
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
30.00%
Implied volatility
25.00%
20.00%
15.00%
10.00%
5.00%
%
70
%
75
%
80
%
85
%
90
%
95
%
10
0%
10
5
11 %
0
11 %
5
12 %
0%
12
5
13 %
0
13 %
5%
14
0
14 %
5
15 %
0%
15
5
16 %
0
17 %
0
18 %
0%
19
0
20 %
0
22 %
5
25 %
0%
30
0
35 %
0%
%
65
%
60
55
50
%
0.00%
Relative strike
FIGURE 2.9 Volatilities with respect to strike price for 3-month options on copper.
Source: Author, LME
pattern is also observed for options on electricity. Note while gold resembled a smile,
this relationship is not quite symmetrical and is sometimes referred to as a ‘smirk’.
Figure 2.9 illustrates the relationship for options on base metals such as copper.
Traditionally, skews in this market have been less pronounced than in other commodity markets.
2.7.5.2
Volatility term structure
Recall that within the BSM option valuation framework, volatility was assumed to be a
constant. In Section 2.7.5.1 it was shown that empirically volatility varied with respect
to option strikes. It can also be seen that volatility varies with respect to time to maturity.
In most financial markets the so-called term structure of volatility is upward sloping
(see for example Schofield, 2017). This suggests that longer-dated options will be valued
using higher volatilities. One reasonable way of interpreting this shape is to suggest that
uncertainty over the ‘at maturity’ outcomes of the underlying asset rise with respect to
time to maturity. However, in many commodity markets volatility displays an inverse
term structure, i.e. volatility decreases with respect to maturity. Consider the following
option on a fictitious asset (Table 2.10). The underlying asset price is 100 and in all
instances the option is struck ATM.
Two points arise from this analysis:
▪ For a given maturity there is a positive relationship between an option premium
and the level of implied volatility.
▪ In all three conditions, longer-dated options will have a higher value than
shorter-dated options.
55
Derivative Valuation
TABLE 2.10 Option premiums in various term structure conditions.
Constant volatility
Positive term structure
Negative term structure
Maturity and vol Premium Maturity and vol Premium Maturity and vol Premium
1m / 15%
3m / 15%
6m / 15%
12m / 15%
1.73
2.98
4.21
5.92
1m / 15%
3m / 16%
6m / 17%
12m / 18%
1.73
3.18
4.77
7.10
1m / 15%
3m / 14%
6m / 13%
12m / 12%
1.73
2.79
3.93
5.53
Source: Author’s calculations
8M
9M
10
M
11
M
12
M
13
M
14
M
15
M
16
M
17
M
18
M
19
M
20
M
21
M
22
M
23
M
24
M
5M
6M
7M
4M
2M
3M
30.00
29.00
28.00
27.00
26.00
25.00
24.00
23.00
22.00
21.00
20.00
1M
Implied volatility (%)
The following figures illustrate the relationship for different commodities:
Recall the forward curves for crude oil in Figure 2.2. That figure illustrated that
the short end of the forward curve tends to be the most volatile. Logically this would
also impact the volatility term structure of options (figure 2.10) and since it is inverted
it suggests a wider dispersion of ‘at maturity’ values for shorter-dated options than
longer-dated options. Since it is plausible to suggest that commodity prices mean revert
then at a simple level, there is less expected variability for longer-dated options. This is
also shown in the forward curve diagram where there is an element of ‘convergence’ for
longer-dated prices. A similar argument could be used to explain the shape of the volatility term structure for natural gas (Figure 2.11). Again, gold is something of an exception.
If one were to accept that it is traded more like a financial asset, then the upward
sloping volatility term structure is not unexpected (Figure 2.12). The figure for copper
indicates that like their observed skew profiles, term structures are not as well defined
(Figure 2.13).
Option maturity
FIGURE 2.10 Term structure for crude oil.
Source: Author, CME group
56
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
50.00
Implied volatility (%)
45.00
40.00
35.00
30.00
25.00
8M
9M
10
M
11
M
12
M
13
M
14
M
15
M
16
M
17
M
18
M
19
M
20
M
21
M
22
M
23
M
24
M
6M
7M
5M
4M
2M
3M
1M
20.00
Option maturity
9M
10
M
11
M
12
M
13
M
14
M
15
M
16
M
17
M
18
M
19
M
20
M
21
M
22
M
23
M
24
M
8M
6M
7M
5M
4M
2M
3M
18.00
17.50
17.00
16.50
16.00
15.50
15.00
14.50
14.00
13.50
13.00
1M
Implied volatility (%)
FIGURE 2.11 Term structure for natural gas.
Source: Author, CME group
Option maturity
FIGURE 2.12 Term structure for gold.
Source: Author, CME group
21.00%
20.50%
20.00%
19.50%
19.00%
1m
3m
6m
FIGURE 2.13 Term structure for copper.
Source: Author, CME group
1y
2y
3y
5y
10y
Derivative Valuation
57
What is the relevance of smiles and skews? First and foremost, options on
commodities should always be valued considering the degree of skew or the term
structure of volatility. But there are more subtle applications. In Chapter 14 a variety
of commodity-structured products will be considered. Typically, these will incorporate
some form of long-dated option. If the option references an asset with a downward
sloping term structures, it will make these structures relatively more attractive for
the investor.
CHAPTER
3
Risk Management Principles
3.1
DEFINING RISK
The use of the phrase ‘risk management’ needs to be clarified. For our purposes, risk
within a financial context can be broken down into five major subcategories:
▪ Market risk – the risk that something that is owned or owed will change in value as
market prices change.
▪ Credit risk – the risk that monies owed will never be repaid.
▪ Business risk – the risk that an entity engages in a business, which they do not fully
understand.
▪ Operational risk – the risk that an entity loses money because of an operational
control weakness.
▪ Legal and documentary risk – the risk that a contract is deemed to be null and void
(e.g. ultra vires), or that a loss is incurred because of a contractual commitment.
Although most institutions would be concerned about market and credit risk, in
reality most catastrophic losses have probably been caused by a lack of understanding of
the business in which the entity has been involved (business risk) or very poor internal
controls (operational risk). In some ways there is an irony in that the risks that are
most commonly at the root of most institutions’ losses cannot be easily hedged. For
example, poor internal control issues are usually a result of cultural issues and a lack of
discipline. For market risks a host of products exist that allow an institution to protect
themselves against an adverse move in prices. Next time there is an announcement of
losses incurred by an institution that are initially blamed on price movements, consider
how it was that the perpetrator was allowed to engage in their activities for so long
without any independent control highlighting the problem.
3.1.1
Subcategories of risk
Market risk
Within each of the individual risk headings it is possible to break down the risk into a
series of subheadings. For example, within market risk one could consider:
58
Risk Management Principles
59
▪ Interest rate risk – the risk that entity loses money from an adverse movement in
interest rates. For example, a company may borrow money on a variable rate basis
and then be faced with higher costs if interest rates subsequently rise. Equally this
exposure also applies to interest earned on cash surpluses, which is often ignored.
▪ Inflation risk – although not always obvious, this may be an exposure for a corporate
whose revenues or expenses are linked to inflation. A corporate may be faced with
a hidden inflation cost if they were to offer a pension scheme to their employees
which is linked to a change in inflation
▪ Foreign exchange rate risk – typically foreign exchange rate risk arises from two
sources. Transactional FX risk arises because of a company’s day-to-day business.
For example, a US importer of goods and services from abroad will have a foreign
currency payable. Translational FX risk arises from expressing a foreign currency
asset or liability in the company’s domestic accounting currency. With respect to
commodities, since they are mostly traded and quoted in US dollars, foreign producers or consumers will also be exposed to FX risk.
▪ Equity risk – this can arise in several different forms from a corporate perspective.
Again, one source of equity risk could be hidden in a company’s defined benefit
pension scheme. Proposed share buybacks and employee share option schemes
may require the company to buy equity at an unfavourable price. Investments in
other publicly quoted companies gives rise to a form of equity investment risk
▪ Liquidity risk – liquidity risk can be thought of as the potential inability of a company to meet its short-term cash requirements. This may arise from the inability
to borrow money from its bankers, or the inability to be able to liquidate assets to
cover any shortfall.
▪ Commodity risk – simply, this is the exposure that a company will face because of
a change in commodity prices. This may be either explicit or as a side product of a
company’s business. For example, a gold producer will be exposed to a fall in the
price of the commodity, which is a direct function of his production. A haulage
company’s exposure to diesel prices is arguably a secondary exposure to their main
line of business. A company that is fully integrated along a particular supply chain
will arguably face offsetting price risks.
Credit risk
The other main risk from the ‘big five’ noted previously that can be hedged is credit risk;
although the use of credit derivatives is outside the scope of this book.
To illustrate the principles let us consider the issues faced by a consumer of base
metals within the automotive sector. It would be possible for them to enter direct
physical hedging contracts with the producers of the metal, but this is not without
difficulty. There would be three main risks that the producers would face. Firstly, there
is the issue of product specification. Hedging directly with say a mine is not necessarily
a good idea, as they will not be producing the metal in its primary form. It is more
likely to be in an intermediate form that needs further work to make it usable by the
consumer. Although a physical hedge may be possible with a smelting company, there
is still no guarantee that they will be able to produce the metal in the exact specification
required. The second issue relates to the counterparty credit risk. If either the mine or
60
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
the smelter suffers any operational problems, they may not be able to meet their future
physical obligations. Finally, there is the issue of price expectations. Consumers of the
metal will be looking to hedge at times when the commodity price is at the bottom of
the price cycle, while producers will be looking to hedge towards the peak of prices.
One of ways that the metal consumers have sought to get around this is through
fixed price component supply. However, this raises the issue of credit risk once again.
If the metal consumer enters into an agreement with a component manufacturer, the
latter will have to decide whether to hedge them to avoid a price mismatch. If the component suppliers choose to hedge directly and are of a poor credit standing, the terms
of their hedge may not be favourable and this may result in poor terms being passed
on to the automotive producer. Alternatively, if the component supplier chooses not to
hedge their raw material costs, they will be impacted if the price of the underlying metal
subsequently rises.
Legal risk
A force majeure provision in a contract means that a party may be relieved from performing some or all their obligations. It is also referred to as an ‘Act of God’ clause.
Typically, they will apply when a specified event occurs (usually taken to be events
that are beyond the control of the affected party), which prevent them from performing
under the terms of the contract. During the COVID-19 pandemic of 2020 instances of
this were seen in the commodities market. For example, both Chinese copper and LNG
buyers faced a drop in demand due to reduced economic activity. However, the Financial Times (2020B) points out ‘LNG sellers complain that China’s use of force majeure
is at least partly motivated by importers’ desire to buy at cheaper spot prices instead of
cargoes imported under long-term contracts.’
3.2
COMMODITY MARKET PARTICIPANTS – THE TIME DIMENSION
When analysing commodities, it is useful to try and develop an understanding of the
likely activity of market participants with respect to time. This is done by analysing
the forward curve, which indicates the clearing price where demand equals supply for
different delivery points in time. In each of the subsequent chapters the price drivers for
each specific commodity are considered, but here we will try and provide a generalised
approach to the issue of participant’s activity.
3.2.1
Short-dated maturities
Typical participants and their related activities in these markets are:
▪ Companies (e.g. banks and commodity trading companies) managing short-term
trading risk.
▪ Investors tracking commodity indices.
▪ Hedge funds and commodity trading advisors using futures to express short-term
views on the market.
Risk Management Principles
3.2.2
61
Medium-dated maturities
These maturities are dominated by the strategic risk management activities of producers
and consumers. Examples of this include oil refinery margin hedging and consumer
hedging such as airlines looking to hedge their jet fuel exposures.
3.2.3
Longer-dated exposures
These maturities will see producers who are selling their production forward, perhaps
as part of the terms of a loan facility where the lending institution is seeking to minimise the volatility of the borrower’s cash flow. If the curve is steeply backwardated
consumers may be tempted to take the opportunity to buy forward at what could be
an attractive level. The other main participants in these maturities are investors buying
OTC structured notes issued by investment banks. As a generalisation these notes may
be constructed as a zero-coupon bond plus a call option.
3.3
HEDGING CORPORATE RISK EXPOSURES
When deciding whether to hedge, a corporate is faced with several possible strategies.
These will be considered from the perspective of an automotive metal consumer to give
the examples a context.
Do not hedge – Here the decision not to hedge may be driven by the fact that the
metal may only make up a small proportion of the finished product and so price
fluctuations will only have a limited impact on margins. It may also be that the
producer can pass on any increase in the price of the underlying metal to the consumers without impacting their own margins. If the underlying price were to fall,
the producer can of course cut the final price, but the price reduction may be as
large as the underlying price falls.
Hedge to guarantee future unit costs – In this instance the manufacturer tries to
lock in much of the future cost of a production run by hedging a significant proportion of the commodity price exposures. This allows the price of the final product
to be fixed and ensures that there is an adequate margin. It also benefits the end
customer, as they are immune to excessive price volatility.
Timing mismatch between raw material cost and final product income –
This situation arises if the manufacturing process is relatively long, and the price
of the final product is linked in part to a commodity price. Having paid for the raw
material, the manufacturer will be exposed to a fall in the price of the commodity
prior to its sale. If the company were able to receive the raw material on a constant
basis and match this with sales of their final production, the timing mismatch may
not be a significant issue. However, in reality companies may receive irregular shipments of the raw material and may also choose to carry large levels of stocks. Thus,
the timing mismatch issue becomes more acute.
62
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Active hedging based on directional views of the market – Many nonfinancial companies are uneasy about developing views on potential commodity
price movements. The result would be that the hedges would only be implemented
based on planned production levels. However, a number of firms have been
willing to take advantage of favourable market conditions to implement hedges in
anticipation of future needs as the companies realise that they will always have a
certain need for a particular commodity.
The other difficult decision is determining the size of the hedge. Hedges may be
put in place of a forecasted purchase of the raw material. If the actual raw materials are
not purchased, then the manufacturer may be faced with a hedge for which there is no
underlying exposure. For this reason, several companies choose to hedge a proportion
of their exposure. Many of the hedges put in place by companies are financially settled.
This allows the hedger to source the metal from a particular supplier who can provide
a specific grade or shape.
3.4
A FRAMEWORK FOR ANALYSING CORPORATE RISK
Very often risk management solutions to a potential market risk are considered at
a micro level, without considering the big picture implications for the corporate.
This approach ignores the fact that a corporate will be ultimately focused on some
high-performance metric such as the share price, profit, or cash flow. As a result,
it would be more appropriate to consider a two-stage approach to the issue – strategic
and tactical. The following framework considers risk management, which develop
ideas presented by Banks and Dunn (2003).
3.4.1
Strategic considerations
At the strategic level, a risk management strategy should consider the following issues
so that the hedging solution can be placed within a particular context:
▪ What are the company’s strategic corporate goals?
▪ What issues are important to the internal and external stakeholders?
▪ What is the universe of financial risks to which the company is exposed? Which of
the financial risks can be hedged?
▪ What is the appropriate strategic performance metric against which the success of
the hedging programme is assessed? (e.g. share price, cash flow)
▪ What is the company’s existing risk management strategy?
▪ What is known about the competitors’ activities?
3.4.2
Tactical considerations
Only after the strategic issues have been addressed should the tactical side of hedging
be considered. Some of the key issues to be addressed include:
Risk Management Principles
63
▪ What is the nature of the specific exposures to which the company is faced?
– How certain are the exposures?
– When should the hedge start? Now? In the future?
What percentage of the underlying exposure should be covered? (i.e. how much
risk are they willing to take?)
– Is there a specific target rate or price that the customer is trying to achieve (e.g.
some form of budget rate)
▪ What is the company’s current view on the market? This could be with respect to:
– Direction of the market
– Timing
– Magnitude
▪ What hedging instruments is the company allowed to use?
▪ Is the company willing to pay a premium fee for protection?
Corporate hedgers may be reluctant to express a view on the future movement of the
market, sometimes claiming that they are not qualified to express a view on the market.
However, it would be fair to say that any view expressed by an investment is just that – an
opinion rather than a guarantee. A more efficient hedge could be constructed by a bank
if the customer were able to express a view on these elements as option based strategies
are normally three dimensional – they extract value from directional movements in the
underlying price, the time to maturity, as well as implied volatility (i.e. the magnitude
of expected price movements).
Reasonably many hedgers may be concerned about the potential outcome of the
hedging strategy. This can be addressed through scenario analysis. By constructing a few
simple ‘what if’ scenarios the potential outcome of the hedged position could be shown.
Indeed, the use of scenarios should ideally feed back into the strategic questions that
are addressed at the start of the process. We argued that all hedging strategies should be
placed within a larger context, as most firms hedge to solve a particular problem, such
as a share price decline or an adverse movement in cash flow. The scenario analysis
should feed back into how the hedged position will influence this key metric. Readers
interested in this approach are referred to Charles Smithson (1998) for an example.
3.5
HEDGING CUSTOMER EXPOSURES
The trading function of a commodity market participant (be it a bank or a trading house)
will be responsible for making decisions on how to manage the bank’s market and credit
risk. These risks arise either from client business or from transactions designed to profit
from a trader’s view as how market prices are expected to evolve. The way in which
these risks are managed will have an influence on market activity and consequently
the market price.
The essence of derivative solutions is that they offer to market participants (corporates, institutional investors, and other financial institutions), the ability to transform
some form of market or credit risk. As a result, the institution offering a particular solution will be taking on some form of risk and will therefore look to offset this in the
market participant. This will also apply to both asset and liability structures.
64
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
It would also be reasonable to suggest that in addition to managing exposures generated by client business, some traders will have discretion to execute trades that take
advantage of an anticipated movement in market rates. Looking at all of these perspectives, a trader’s ‘book’ would therefore comprise a portfolio of different exposures that
they will manage on an ongoing basis.
Throughout the book we will analyse how commodity exposures are created and
hedged using the suite of derivative products:
▪ Futures and forwards
▪ Swaps
▪ Options (vanilla and exotic)
3.5.1
Forward risk management
If a client were to enter an exchange traded futures contract, this does not result in any
credit risk for the bank, as the client’s counterparty is the exchange’s clearing house.
Any change in the value of the future is managed by the clearing house, through the
collection of collateral (margin calls) from the customer.
If a client executes a forward deal, the bank is now acting as a principal to the trade
and as such this generates both credit and market risk. Given the way in which a forward
price is calculated (i.e. spot price plus net carry) the bank has an exposure to a movement
in all of these components. By taking an equal and offsetting position the bank can
manage these risks. This concept is illustrated in the chapter on gold, where the pricing
and risk of a forward deal is considered in detail.
3.5.2
Swap risk management
Hedging of swap risks can be more complex but it is possible to make some general
observations. Again, the simplest hedge for the bank is an equal and opposite transaction with another market participant. However, if even the particular swap market
is exceptionally liquid, this may be difficult to achieve. As a result, many banks will
select instruments that are a close proxy for the exposure and run an element of market
risk, usually within defined parameters. This type of market risk is referred to as basis
risk and describes a situation where the value of the exposure does not change by the
same amount as the hedging instrument. For example, in certain types of energy swap,
based on one or more of the refined products (e.g. jet fuel), the traders may convert the
exposure to a crude oil equivalent and manage the risk using exchange traded futures.
3.5.3
Option risk management
When a bank executes an option transaction it is faced with several different market
risks. The market risks they look to manage relate to changes in:
▪ The underlying price.
▪ The strike price.
▪ Time to maturity.
Risk Management Principles
65
▪ The cost of carrying an underlying hedge.
▪ The implied volatility of the underlying asset.
Each one of the listed market risks has an associated ‘Greek’ measure that is used
by the banks to measure and manage a particular exposure. These Greeks and how they
are used to manage market risk were analysed in Chapter 2.
3.5.4
Correlation risk management
One additional market risk worthy of mention is that of correlation risk. Correlation
measures the tendency of variables to move in the same direction. Within the context
of this book the focus will be on price correlation, although it is acknowledged that
alternative correlation measures exist elsewhere within finance (e.g. default correlation
within the credit world).
Correlation risk may arise in several ways:
▪ From options that require it as a pricing input (e.g. spread and basket options).
▪ The price relationships that commodities have with each other (e.g. platinum and
palladium).
▪ In the price relationships commodities have with financial assets such as bonds or
equities.
Similar to volatility, correlation could be estimated from historical data or be
implied from current market prices. However, correlation measures tend to be unstable
in volatile markets.
3.5.5
Case study: Managing market and credit – the collapse of Japan
Airlines
One example of credit risk within the context of commodities was the collapse of Japan
Airlines in 2010 with debts of YEN 2,322 billion. At the time this was Japan’s fourth
largest corporate failure with the government pledging YEN 900 billion to keep the airline operational. The airline had hedged at least 78% of its fuel consumption but it is
likely that some of this was done in 2008 and 2009 when energy prices were trading at
all-time highs. The outstanding fuel hedges were estimated at USD 440 million, which
was probably made up of forwards, futures, and swaps. To illustrate the nature of the
counterparty risk, consider the following simplified example.
In Figure 3.1 the airline establishes a commercial contract with a jet fuel supplier
for delivery of the required volume of the contract. They agree between themselves that
the contract price will reference some agreed index. To hedge this risk the airline enters
a swap contract where they will receive the same index price for jet fuel that they are
paying on the underlying contract. In exchange, they will pay to the bank a fixed price.
The net result is that the airline is immunised against movements in the market price of
jet fuel, but on a net basis will pay a fixed price for their fuel. The swap has market risk
in that they will lose money if the price of jet fuel were to rise. As a result, they hedge
66
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Fixed price
Bank
Airline
Market
Price of
Jet Fuel
Physical
Jet Fuel
Jet Fuel
Supplier
Market
Price of
Jet Fuel
Long
crude oil
futures
Seller
FIGURE 3.1 Jet fuel swap.
themselves with a long futures position in crude oil. This is not without risk as they are
assuming that the price of crude oil will move in the same direction as jet fuel. Airline
hedging strategies are considered in more detail in Chapter six.
From the bank’s perspective credit risk only exists where the airline defaults and
the derivative is in-the-money (ITM). Suppose that the fixed price on the swap had been
set at USD 800.00 per metric tonne. If the market price of jet fuel were to fall, then
at a very simple level, the swap would have value to the bank as they are paying out
less than they are receiving. If the airline were to default, then market practice is for
the derivative to be valued with any positive ‘mark to market’ (i.e. market value) being
settled between the two entities. If the swap were ITM from the bank’s perspective,
then they would become an unsecured creditor and would have a claim on the airline.
If the swap were OTM for the bank they would be required to settle their debts with
the airline’s bankruptcy practitioner. In the case where the swap is OTM for the bank,
they do not have any credit risk. Credit risk only exists when an entity is owed money;
money owed to the airline is a debt that must be paid. Readers interested in a more
detailed explanation of derivative credit risk are referred to Gregory (2020).
Since the associated swap cash flows will no longer take place, the bank is now left
with an underlying hedge position. To unwind this hedge the bank will be required to
sell futures back into the market, which in the case of Japan Airlines caused the market
to fall in the short term.
3.6
TRADING RISK MANAGEMENT
What is trading? Although there is no single universal definition of the term, it can be
characterised as the buying and selling of a commodity with profit as the main motivation. Rather than having to buy and sell the commodity or a derivative because of some
underlying operational need or the desire to offset some economic exposure, trading is
decidedly ‘view driven’.
Risk Management Principles
67
There are a whole series of strategies that fall within this category and as such the
following highlights a few key themes. Throughout the text other examples will be illustrated within a particular context.
3.6.1
Spot trading strategies
A trader may decide to take an outright view on the movement of an asset. An outright
deal is a single buy or sell strategy rather than some combination of deals. For example,
say a trader is looking at an investment period of six months, but is not sure whether
he should buy or sell. The decision is not as simple as suggesting that if he expects the
price to rise then he should buy and vice versa. The decision is a function of the investment horizon and how the trader thinks that spot prices will evolve relative to initial
forward price. Let us assume that gold is trading in the spot market at USD 1,425.40
per troy ounce while the six-month forward is priced at USD 1,449.10 (these are the
same values as those presented in chapter two). The forward price is not a predictor of
the future spot prices but should be used as a breakeven value against which all investment decisions should be judged. This is because a spot purchase which is held for a
given period, will have a value equal to the forward price. This gives the trader one of
three possibilities:
▪ If the trader believed that the actual spot price in six months’ time was going to
be greater than the current forward price, then they should buy the gold, borrow
money to finance its purchase, and lend out the gold to earn the lease rate. If their
view materialises, they will be able to sell the gold at the end of the period at a price
that will be greater than the cost of carrying the metal for the period.
▪ If the trader believed that the actual spot price in six months was going to be
less than the current forward rate, then a selling position would be appropriate.
They should sell the gold and place the proceeds on deposit to earn interest for
six months. To ensure that the short position is covered, the bank would borrow
the gold and pay a leasing fee. At the end of the six-month period, the gold is
repurchased at the prevailing spot price and delivered to the lender of the metal.
The net result is that the cash flows received exceed the cash flows paid.
▪ If the trader believed that the actual spot price in six months was going to be equal
to the current forward rate, then they would be indifferent between buying and
selling the gold.
Note that in the second scenario it would be possible for the trader to believe that
the price of gold was going to rise, albeit to a level less than the current forward rate,
and this would still indicate a selling strategy. The key therefore is not whether the spot
price will rise or fall but where it will be in relation to the forward rate.
3.6.2
Forward trading strategies
One obvious motivation for executing a forward trade is the fact that the trader does
not have to take physical delivery of the underlying asset. In this sense it may be viewed
as a more cash efficient manner of expressing a view on the market. Similar to the
68
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
spot market, forward starting deals could be used to express a view on a directional
movement in the underlying price.
However, it may be more common to use forward transactions to exploit relative
price movements. These would be executed as a pair of trades and may include:
▪ Taking a view that contracts trading with two different maturities may be mispriced
relative to each other. Here the trader buys the contract considered to be underpriced and sells the contract that is overpriced. This strategy is based on how the
shape of the forward curve is likely to move. In essence, curves could move in one
of two ways. Firstly, the slope between any two maturities may increase or decrease.
So, if a trader expected the slope of the forward curve to steepen the classic trade
would be to sell the short-dated future and buy a longer-dated future. Secondly, the
curvature of the forward curve could change. If the trader expected the curve to
become more concave, then they would sell a short- and long-dated future and buy
a medium-dated future.
▪ Taking a view that a historic relationship between two related prices may change.
For example, a trader may identify a seasonal relationship between prices for delivery in winter and spring they believe will change. This is explored in greater detail
in the Amaranth Advisors case study, which appears later in this chapter.
▪ A trader may initiate a transaction that exploits a differential between two different yet related commodities. For example, popular spread trades in the commodity
markets include:
– Crack spreads (e.g. gasoline or heating oil vs. crude oil).
– Spark spreads (e.g. natural gas vs. power).
– Base metal spreads (e.g. copper vs. aluminium).
– Precious metal spreads (e.g. gold vs. silver or platinum vs. palladium).
3.6.3
Single period physically settled ‘swaps’
These are constructed as a simultaneous combination of a spot and a forward transaction. It is a temporary exchange of an asset for cash, which can be used for either:
▪ Borrowing money at an attractive rate.
▪ Obtaining temporary supplies of a commodity.
A fully worked example of this type of structure is documented in the chapter
on gold.
3.6.4
Single or multi-period financially settled swaps
A financial swap is either a single or multi-period exchange of cash flows that does not
involve the physical movement of the commodity. A single-period swap will be economically equivalent to a cash-settled forward. From a trading perspective swaps could be
used in a similar manner to forwards.
Risk Management Principles
3.6.5
69
Option based trades – trading volatility
Recall from Chapter 2 that options are priced using several variables. The variables are
agreed with the counterparty (e.g. maturity, strike) while other variables are easily discernible from market sources (e.g. the underlying price and interest rates). However,
there is one variable – implied volatility – that although may be discernible from market
sources is the source of a number of trading opportunities.
Implied volatility is a measure of how volatile the market expects the asset to be
until the option’s expiration. Since it attempts to quantify something in the future, it is
nothing more than an educated guess – a perception of how risky the asset is expected
to be. Market price screens do exist for implied volatilities, but these simply represent a consensus or equilibrium. However, it needs to be grounded in reality and so
should bear some resemblance to how volatile the asset has been over a particular period
(e.g. historical volatility). Expressed as an option ‘Greek’ implied volatility is referred to
as ‘vega’.
Accordingly, traders talk about buying and selling (implied) volatility as if it were a
commodity. Recall from the pricing chapter 2 that an option buyer benefits from a rise
in implied volatility, while sellers benefit from a fall. In some respects trading volatility
is simple; if an option is bought, a premium is paid and a subsequent rise in implied
volatility allows the buyer to take a reversing position in the market (‘sell the option
back to the market’) to receive a higher premium. However, this strategy is not without
risk as the option position leaves a trader with a number of other market risks (e.g.
changes in the spot price, passage of time) that need to be managed. The following
example illustrates how volatility could be traded and the associated risks.
Trading of volatility is achieved by trading buying or selling options to achieve a
particular vega exposure, while neutralising the position’s exposure to movements in
the underlying price. The typical strategies for achieving this are:
Selling volatility (trader believes that implied volatility will fall)
i. Sell an at-the-money (ATM) call, sell an ATM put, same strike, same maturity (known as a short straddle)
ii. Sell a call option and delta hedge
iii. Sell a put option and delta hedge
Buying volatility (trader believes that implied volatility will rise)
i. Buy an ATM call, buy an ATM put, same strike, same maturity (known as
a long straddle)
ii. Buy a call option and delta hedge
iii. Buy a put option and delta hedge
Initially it can be confusing to grasp how volatility could be traded using the
single option strategy, but this is usually because the option’s exposure is viewed along
a single dimension. It would be more accurate to say that when executing option
transactions, a trader must think in two or three dimensions, namely volatility (actual
and implied) and direction. The sale of either a call or a put will result in a negative
vega exposure while the delta exposure is negative for the sold call and positive for
the sold put. This directional exposure is neutralised by buying the underlying (for the
call) and selling the underlying (for the put) a technique referred to as delta hedging.
70
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Let us assume a bank has sold a three-month out-of-the-money (OTM) call option
on gold on a notional amount of 10,000 troy ounces with a strike rate of USD 1,450,
while spot is USD 1,425.40. With implied volatility at 15% the premium will be USD
31.66 per troy ounce. To hedge the delta exposure the trader decides to use the spot
market, which is currently trading at a price of USD 1,425.40 per troy ounce.
The delta of this OTM option is −0.4223, which means that if the price of the underlying were to change by a small amount, the premium on the option would change by
42.23% of the amount. Since a rise in the underlying price would cause the option to lose
money, the trader decides to hedge this exposure by buying the underlying in the proportion dictated by the delta value. Since the option was written on a notional of 10,000 troy
ounces the trader must buy 42.23% of this amount, namely 4,223 troy ounces. Therefore, if there is a small increase in the price of gold, the losses on the option should
approximately be offset by the profits on the delta hedge.
All other things being equal – mainly that the underlying price does not
change – holding this position to expiry will yield a profit to the trader. The passage of
time will reap a profit to the trader through the accretion of the option’s value by the
theta effect. If this assumption is relaxed the main risk to the trader is to large changes
in the underlying price. At first glance this may seem counterintuitive as the delta
hedge was designed to ensure that the trader’s directional exposure was neutralised.
However, although delta neutralises the trader against small changes in the underlying
price, he is still exposed to large changes in the underlying price.
Let us say that shortly after executing the trade, the price of gold rises by USD 5.00
per troy ounce to reach a level of USD 1,430.40. The option position is now valued at
USD 33.82 per troy ounce, a loss to the trader of USD 2.16 per troy ounce or USD 21,600
on the option position. However, the trader will gain USD 5.00 per troy ounce on the
underlying hedge. Since the delta hedge was for 4,223 troy ounces, the profit would
be USD 21,115, which is insufficient to cover the loss on the option position, resulting
in an overall mark-to-market loss of USD 485.00. At this point the trader is faced with
a difficult decision. If the price of the underlying continues to rise, the losses on the
option, which move in a non-linear fashion, will increase faster than the profits on the
future, which change in a linear manner. As a result, he decides to rehedge himself at
the new higher level where the delta on the option is now –44%. This means to be delta
neutral he needs to buy a total of 4,400 troy ounces of gold. Since he already owns 4,223
troy ounces, he only needs to buy the difference (177 troy ounces), albeit at a new higher
price of USD 1,430.40 per troy ounce.
After rehedging the trader then sees that the market retraces falling back to its
opening price of USD 1,425.40. On a mark-to-market (MTM) basis, the option position
gains in value by USD 2.16 or USD 21,600 for the position. However, the physical hedge
position in gold now loses USD 5.00 per troy ounce to give a hedge loss of USD 22,000
(4,400 ounces x USD 5.00). His net loss from this price fall is therefore USD 400.00. His
total losses as a result of both price movements are therefore USD 885.00 (USD 485.00
+ USD 400.00).
If there were no further price movements during the trading day, the trader will
have:
▪ Broken even on the option position (note the analysis is done on an intraday basis
so theta is not an issue),
Risk Management Principles
71
▪ Broken even on the initial hedge position (4,223 troy ounces bought at USD
1,425.40),
▪ Lost money because of the rebalancing of the portfolio (an extra 177 troy ounces
bought at USD 1,430.40 and revalued at the close of business at USD 1,425.40 to
give a total loss of USD 885.00).
What would have happened if the price had first fallen and then subsequently risen?
If the price of the underlying had fallen the delta on the option would also have fallen,
indicating to the trader that they were overhedged. He would then have to sell part of
the hedge at a lower price. However, if the price were subsequently to rise then the
hedge would have to be repurchased at a higher price. The result is therefore the same,
when rebalancing the trader will be buying the hedge at a higher price and selling it at
a lower price.
This is the primary risk of selling options on a delta-hedged basis into a volatile
market. The rate at which the option changes in value is different to that of the hedge.
If the price of the underlying were to become more volatile the delta of the option would
change, indicating to the option trader that the delta hedge needs to be adjusted. The
magnitude of the change in delta as a result of a move in the underlying price will dictate the size of the rebalancing. With a short option position in a volatile market this will
result in a mark to market loss. The rate of change of delta with respect to the underlying
price is called gamma. Gamma is always expressed in relation to a range of price movements. So, in the above example the gamma for a USD 5.00 movement in the price of
gold is 1.77% (44%–42.23%). Intuitively, gamma can be thought of as:
▪ The trader’s exposure to significant changes in the underlying (since delta only covers him for small changes).
▪ A trader’s exposure to movements in actual as opposed to implied volatility.
Additionally, some traders think of gamma as the speed with which their profit and
loss changes with respect to a change in the underlying price. The larger the gamma,
the greater the difference between two delta values for a given change in the underlying
price and the larger the required adjustment to the delta hedge. Hence an option with
a large gamma value will generate large changes to the trader’s profit or loss.
If the trader had bought options and delta-hedged under the conditions noted
above, then he would have enjoyed a profit. However, there is a cost to this, as the
options will decay in value over time. Therefore, option traders sometimes describe
theta as “gamma rent”; it is the ‘price’ one must pay when buying volatility or the
‘reward’ one receives from selling volatility.
An option trader who buys volatility is hoping that:
▪ Implied volatility will increase, leading to an increase in the value of their option
position.
▪ The market will experience more (actual) volatility.
▪ Some combination of the two.
If they decide to take a view on the actual volatility of the market, they hope to make
more in delta hedging profits than they lose in time decay.
72
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
An option trader who sells volatility is hoping that:
▪ Implied volatility will decrease, leading to a decrease in the value of their option
position.
▪ The market will experience less (actual) volatility.
▪ Some combination of the two.
If they decide to take a view on the actual volatility of the market, they hope to make
more from time decay than they lose in delta hedging activities. The preceding example
was based on intraday movements in order to illustrate that the source of profit or loss
was from the rebalancing of the hedge position. However, the initial theta of the option
was USD 0.2266 and so as a result of the passage of time, the trader will make USD
2,266, which is greater than the delta hedging loss.
As the underlying price evolves traders may adjust their implied volatility quotations accordingly. One simple process to do this is as follows. By convention implied
volatility numbers tend to be quoted as percentage per annum. It is possible to convert
this to a daily value by dividing by the square root of the number of business days in
the year (e.g. 250). An annual volatility of 15% is therefore equivalent to a daily figure
of 0.95%. The trader would expect the underlying price to trade approximately within
plus or minus of this range. Using our previous opening futures price of USD 1,425.40
this would imply a range of values from USD 1,411.86 to USD 1,438.94.
3.6.6
Case study: Amaranth Advisors and the US natural gas market
‘We thought about pulling the trigger and taking the loss . . . we had many discussions about it. We figured we could get out for maybe a billion dollars. But
we decided to ride it out and see if the market would come around’.
—Amaranth trader (United States Senate, 2007)
Shortly before its collapse in the fourth quarter of 2006, Amaranth Advisors LLC
managed a hedge fund valued at about USD 9.2 billion. The fund lost about USD
6 billion over a two- to three-week period because of losses on its US natural gas
futures trades.
US natural gas prices tend to display an element of cyclicality. That is, the price of
natural gas for winter delivery is greater than for summer delivery so as a result, the
forward curve for natural gas is humped in nature (see Figure 3.2).
In the gas market so-called ‘spread trades’ are a common transaction. A spread
trade involves taking a long position in a future of one maturity, and a short position in
another maturity. A company that distributes natural gas to an end consumer may buy
natural gas for summer delivery and simultaneously sell the gas for winter delivery.
If the cost of storing the gas for this period is less than the spread between the two
prices, then this activity will be profitable for the company.
Of course, for every buyer there must be a seller who will be willing to take the
opposite side of the trade, i.e. sell the gas for summer delivery and buy for winter delivery. A participant’s motive to do this trade could be driven by the view that winter prices
are expected to be relatively higher than summer prices. Note that the spread between
the prices is important rather than the absolute level of prices.
73
Risk Management Principles
Price of natural gas
(USD / MMBtu)
Peaks are winter months
Troughs are summer months
Contract
maturity
FIGURE 3.2 Hypothetical forward curve for US natural gas.
Source: Author
Amaranth’s fundamental view was that winter natural gas prices would be
much higher than summer natural gas prices. They believed this would be a result of
increasing domestic demand for natural gas, expected supply shortages, delivery bottlenecks, and weather-related disruptions. Although the fund had a number of different
positions, it expressed this view through two major trading strategies during 2006:
▪ Selling November 2006 futures and buying January 2007 futures.
▪ Buying March 2007 futures and selling April 2007 futures.
Each spread trade is traded as a single transaction with a single price.
The second trade has been termed the ‘widow maker’ as this particular price spread
is very volatile. This is because March is the last month of the winter heating period
when gas supplies are low and gas is being withdrawn from storage, while April is the
first summer month when gas storage facilities start to be refilled. Amaranth’s view was
that March prices would rise relative to April, i.e. that the spread between the two prices
would increase.
In late August 2006, market prices started to move against Amaranth. Before then
the spread had been trading between USD 1.50 and USD 2.50. In the physical natural gas
markets, there had been little hurricane activity to disrupt production and the amount
of natural gas in storage was very high. Thus, the prices for the two key spread trades
started to fall, i.e. the price of natural gas for winter delivery fell relative to that for
summer delivery. By mid-September, the spread had fallen to USD 0.50 per MMBtu.
Considering all their positions, Amaranth held as many as 100,000 natural gas
futures representing one trillion cubic feet of natural gas. This was equivalent to about
5% of natural gas used in the US in a year. At times Amaranth also controlled up to 40%
of all open positions on the futures exchange NYMEX for the winter month expiries of
October 2006–March 2007.
74
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Given the size of their positions they were required to post increasing amounts
of margin to the exchange. These margins eventually reached USD 3 billion. By
mid-September, the fund was forced to sell its positions to its prime broker and another
hedge fund, and liquidated the fund’s remaining holdings.
3.6.7
Case study: Metallgesellschaft
Metallgesellschaft AG, or MG, is a traditional German metal company, who in the early
1990s evolved into a provider of energy risk management services. They had several
subsidiaries within its energy group, with MG Refining and Marketing Inc. (MGRM)
in charge of refining and marketing petroleum products in the US In December 1993,
it was revealed that MGRM had incurred losses of approximately USD 1.5 billion. The
basic problem lay in a mismatch between the maturities of its commitments to deliver
oil and oil products (heating oil and gasoline) and the derivatives used to hedge those
commitments. The US arm of the group had undergone a rapid expansion of its business, through various acquisitions and had developed an ambitious marketing plan,
which resulted in sales of approximately 300,000 barrels per day. In an attempt to win
business MG negotiated contracts as long as five to ten years with individual service
stations around the US, typically at fixed prices. Estimates suggested that they had built
up long-term commitments to deliver up to 160 million barrels over a five-year period.
This strategy left the group with a large exposure. Since they had contracted to sell
oil and oil products in the future at a pre-agreed fixed price, unless hedged the company
would have to buy oil to meet these commitments at the then prevailing market price.
If the oil price were to move up it would result in large losses.
MG chose to hedge this risk by buying exchange traded three-month futures contracts. This strategy was adopted because liquidity is greatest in the shorter-dated maturities even though it introduced a maturity mismatch. By buying futures MG believed
it had implemented a good offsetting hedge position in relation to its fixed price client
contracts. If oil prices were to increase the client contract would be unprofitable, but
the futures hedge would show a profit (and vice versa). In order to maintain the hedge
over the term of the client contracts, MG was obliged to roll over the futures exposure
at each contract expiry.
Initially the prevailing market conditions were such that futures prices tended to
increase as they approached maturity, i.e. the market was in backwardation. This meant
that MG was able to make profits by selling the expiring futures and buying new lower
priced futures to replace them. The position could be reopened by purchasing contracts
for the next delivery period at a cheaper cost of C, again in the expectation that price
would rise.
The traders at NYMEX were aware that MG would have to execute a substantial
number of rollovers each month and started to use this information to increase the price
of the futures contracts they traded with the company. Additionally, in November 1993
OPEC decided not to cut output, which caused prices to fall sharply with the price of
benchmark crude falling to below USD 14.50 a barrel from a high of USD 22.00 mid-year.
As a result, the forward curve fell and moved into contango.
MG’s trading strategy was impacted in two ways. A fall in the price of crude oil
resulted in extra margin calls of USD 200 million on MG’s long futures position. This
Risk Management Principles
75
was a daily drain on their cash flow that was not matched by income from the underlying client contracts. In addition the firm was now incurring a cost to roll the futures
position due to the upward sloping nature of the forward curve; they were now being
forced to close out their futures positions at a lower price than their initial purchase. At
one stage MG was estimated to be losing about USD 30 million every time it rolled over
the futures contracts.
CHAPTER
4
Gold
4.1
THE MARKET FOR GOLD
‘Boys, by God, I believe I’ve found a gold mine.’
—James Marshall
Along California’s Highway 49, tucked away in a beautiful valley in the Sierra
Nevada foothills is the tiny town of Coloma. Running through the centre of the village
is a fork of the American River where on 24 January 1848; James Marshall found some
gold flakes in the streambed, sparking one of history’s largest human migrations.
What are precious metals?
Arguably the most recognizable precious metals are gold and silver. Another important segment market comprises of the Platinum Group Metals (PGMs). Of the PGMs
there are two actively traded metals, platinum and palladium, although the category
includes other non-traded metals such as:
▪ Rhodium
▪ Ruthenium
▪ Iridium
▪ Osmium
4.1.1
Physical Supply Chain
Gold is extracted from open pits or underground mines by blasting or digging. The economics of mining are such that deep level mines extract gold at purities of less than 10
grammes of gold per tonne of dirt (or ore), while for open pit operations the figure can
be below one gramme per tonne. However, at this stage in its production it will contain
several impurities and will not be in any usable form. It is first milled to release the metal
before being refined and partly purified on site to produce doré gold bars. These bars
will then be sent to refining operations dedicated to produce metal in a form that can
be used by a variety of industries. The resulting output from the refiners may be in the
form of bullion bars or ‘blank’ jewellery. Fabricators will then take the refined product
to make jewellery or will build it into other products such as electronics, which benefit
from gold’s physical properties, such as resistance to corrosion, thermal and electrical
conductivity.
76
77
Gold
4.1.2
Intermediaries
Outside of the physical supply chain sits two important entities: intermediaries (such as
investment banks and trading houses) and central banks. Traditionally central bankers
have been thought of as buyers of gold for reserve asset purposes. However, increasing pressures on central banks to make more effective use of their reserves resulted
in them being net sellers for prolonged periods. These periods usually coincided with
times when the gold price was relatively low. Intermediaries such as investment banks
or trading houses will offer a variety of services that include:
▪ Helping mining companies (producers) manage their exposure to a change in market prices.
▪ Lending metal to refiners or fabricators to finance ‘in process’ inventories.
▪ Borrowing gold from central banks.
▪ Helping central banks manage their gold holdings.
▪ Acting as an intermediary between the main players in the production lifecycle.
▪ Trading with other financial institutions (e.g. hedge funds).
From this brief description of the various entities within the supply chain it is possible to envisage a simple supply and demand model. Without considering the relative
size of each component, the fundamental supply of gold comes from three main sources:
▪ New mine production,
▪ Existing central bank holdings,
▪ Recycling of scrap metal (the ‘urban mine’).
The fundamental demand for the metal is driven primarily by:
▪ Jewellery.
▪ Physical investment (e.g. coins and bars).
▪ Industrial requirements.
Mining
Scrap
Jewellery
Flow Supply
Flow
Demand
Central
Banks
Physical
Investment
FIGURE 4.1 Overview of gold market.
Industrial
78
4.1.3
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The London Gold Market
Although gold is traded in a variety of different locations, London is the largest over the
counter (OTC) bullion market in the world. Gold futures can be traded on various global
exchanges such as the CME Group and the London Metal Exchange. The pre-eminence
of London as the centre of OTC gold trading has led to the development of certain market conventions that are now accepted globally. There are three key professional trade
bodies:
▪ London Bullion Markets Association (LBMA) – the LBMA acts as a
standard-setting body for the global wholesale market for precious metals. The
association’s membership encompasses all participants along the physical supply
chain.
▪ London Platinum and Palladium Market (LPPM) – perform a similar role to
the LBMA but for the PGM market.
▪ London Precious Metals Clearing Limited (LPMCL) – clearing relates to all
the necessary operational activities that take place after trade execution but before
settlement. Settlement is the final process that sees the final exchange of cash for the
underlying asset. According to the LBMA’s website (www.lbma.org.uk),‘The clearing system is operated by the London Precious Metal Clearing (LPMCL) which is
owned and managed by the five banks: HSBC, ICBC Standard Bank, JP Morgan,
Scotiabank and UBS. They utilise the unallocated gold and silver, in accounts they
maintain between each other, not only for settlement of mutual trades but also for
trades on behalf of their third-party counterparties worldwide for whom they provide clearing facilities.’
In terms of market size, statistics produced by the London Bullion Market Association (LBMA) for August 2020 show a daily turnover of about USD 70 billion. To
place this into some context the daily turnover of the foreign exchange market is USD
6.5 trillion1 .
Since metals may be of different purities an element of standardisation is still important even in OTC markets. Counterparties trading on a ‘loco London’ basis (literally
location London) will have an assurance that the metal traded will meet the criteria
necessary for ‘London Good Delivery’. This ensures that the different bars are fungible
within the market and will be delivered to a London based vault nominated by the seller.
The technical name for the shape of these bars is a trapezoid prism.
London Good Delivery means that a bar must conform to the following standards:
Weight
▪ Minimum gold content: 350 fine ounces (approximately 10.9 kilograms).
▪ Maximum gold content: 430 fine ounces (approximately 13.4 kilograms).
▪ The gross weight of a bar should be expressed in ounces troy, in multiples of 0.025,
rounded down to the nearest 0.025 of an ounce troy.
1
FX data sourced from BIS.org. Turnover figure is for 2019.
79
Gold
Fineness
▪ The minimum acceptable fineness is 995 parts per thousand fine gold.
Marks
▪ Serial number.
▪ Assay stamp of acceptable refiner.
▪ Year of manufacture (expressed in four digits).
▪ Marks should be stamped on the larger of the two main surfaces of the bar.
As an extension of these market standards there is also a ‘loco London’ price. Settlement for the gold will take place two good London business days after trade date. A deal
transacted on Wednesday, 15 January 2020 will settle on Friday, 17 January 2020. Loco
London is the default basis for settlement for gold in the same way that New York is for
the purchase or sale of US dollars. If an entity wishes to settle a transaction at another
location or for a different level of purity, an adjustment to this benchmark price is normally made.
Reference is often made to the purity of gold, which can be expressed in different ways. When using carats, the purity of gold is measured on a scale of 1 to 24. For
example, gold described as being 18 carats is equal to 18/24th of 1000 parts, i.e. 750
fineness. However, the wholesale financial markets very rarely refer to carats. London
Good Delivery Bars are a minimum 995/1000 fineness (or ‘two nines five’ in the market
parlance).
London Good Delivery Bars represent bars that have a gold content of between 350
and 430 fine ounces although most bars are generally close to 400 ounces (12.5 kg/27
pounds). Delivery will usually take place at the vault of a clearing member of the LPMCL
either on an allocated or unallocated basis. Holding gold on an unallocated basis means
the metal is held in a vault in common with other holdings and the customer has a general entitlement. Holding gold on an unallocated basis incurs only minimal storage costs
but reduces the holder to being an unsecured creditor. That is, the owner has a claim on
the bank where the gold is held but does not have title to specific bars. If gold is held on
an allocated basis the metal is physically segregated from other customer holdings with
detailed records being kept. Segregating the holding in an allocated account increases
the degree of security as the holder is now secured, but will incur substantially higher
charges.
Traditionally, most of the market trading was done on an unallocated basis. The
custodian banks would hold this gold on their balance sheets, but a change in regulations meant that this approach required them to set aside capital to support the business.
Since bank capital will incur a cost, institutions offering this service have tried to move
their client’s gold holding to an allocated basis, which would take them off the balance
sheet (Financial Times, 2014a).
4.1.4
The LBMA gold price
Around the time of the original gold fixing in 1919, the Bank of England’s governor
envisaged ‘an open market for gold in which not only every seller would know that he
would receive the highest price the world could pay but also every buyer would know
80
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
that he would get his gold as cheaply as the world could supply it’ (Financial Times,
2014b).
Like all traded assets, price is determined by the interaction of demand and supply.
However, as is common with many markets, it is convenient to quote a benchmark
figure to facilitate the settlement of a variety of transactions. There are several reasons
why the market needs a benchmark:
▪ Certain institutions such as central banks may have a legal requirement to trade
with complete transparency and so may decide to execute a transaction at the day’s
auction price.
▪ Industrial contracts may reference the benchmark to avoid conflicts of interest that
may arise from an arbitrary choice of settlement price.
▪ Derivative contracts such as options can use the benchmark to avoid disputes over
whether an option will be exercised or not.
Since 1919 the value of gold has been ‘fixed’ twice daily apart from a 15-year period
between 1939 and 1954. However, the use of the word ‘fixing’ has fallen into disrepute
following several scandals relating to other benchmarks such as LIBOR. As part of a
move to reform market practices, between 2014 and 2015 the old-style fixings for gold,
silver, and the PGMs were transferred to independent third party administrators.2
It is now more common to refer to an auction process for gold that generates a
snapshot benchmark value termed the ‘LBMA gold price’ at two predetermined times
during the day (10:30 a.m. and 3:00 p.m.).
There are two main categories of participant in the precious metals auction. A direct
participant will manage their orders in the auction via the administrator’s electronic
platform. Their trades at the end of the auction are settled with other direct participants or can be centrally cleared via the Intercontinental Exchange (ICE). An indirect
participant is a client of a direct participant who will settle any transactions with their
direct participant. At the time of writing there are currently 12 direct participants.
The auction works in the following manner. The benchmark administrators will
first propose a price to the participants. Based on this suggested price the participants
have 30 seconds to submit their buying and selling volumes, which are fully executable
if the auction is ‘in balance’. At the end of the 30 seconds the administrator’s system
will calculate the difference between the buying and the selling volumes to see if there
is a significant imbalance. If the imbalance is greater than 10,000 ounces, then the price
is adjusted based on the direction of the imbalance and then the process is repeated. If
the imbalance is less than 10,000 ounces, then all the volumes will trade at that price.
Once the volume has been traded at this final auction price, it is then declared to be the
USD benchmark price. The administrators will then convert the benchmark price into
a variety of other currencies.
Consider the following hypothetical auction example (Table 4.1).
Suppose the initial suggested price in round 1 drew more buying volume (bids)
than selling volume with the difference being 50,356 ounces. Since this was outside
2
At the time of writing the LBMA gold price is managed by ICE benchmark administration.
81
Gold
TABLE 4.1 Hypothetical example of gold auction.
Round
Price
Bid volume
Ask volume
1
2
3
4
5
1329.15
1329.55
1330.15
1329.90
1329.75
80,579
80,579
80,579
80,579
80,579
30,223
35,223
114,223
94,223
85,223
the 10,000-ounce imbalance threshold, then the price was increased for the second
round. Although the second round attracted more sellers, the imbalance persisted and
so the price increased for a third time. However, the increase in price for the third
round attracted substantially more sellers than buyers and so the imbalance flipped.
The price was then reduced for the fourth and fifth round when the price finally set at
USD 1,329.75 and a demand/supply difference of 4,644 ounces.
In addition to the spot market, two other key parts of the gold market are outright
forwards and swaps. Until 2015 the LBMA operated a fixing to set what was referred to
as ‘GOFO’ (gold forward offered rate). However, this fixing has now been discontinued
although the concept may still be used in wholesale transactions. However, instead of
quoting the forward price in outright USD terms the market makers quote the GOFO
rate as the difference between the spot and forward USD prices expressed as a percent
per annum. These transactions will be considered later.
4.2
GOLD PRICE DRIVERS
‘Gold gets dug out of the ground in Africa or some place. Then we melt it down,
dig another hole, bury it again and pay people to stand around guarding it.
Anyone watching from Mars would be scratching their head’.
—Warren Buffet
‘There is one unique feature of gold that is possessed by no other investment
and which ensures it is an investment of choice for many: its place in the
human psyche. It has been a global monetary unit for more than 2,500 years
while also having significance in language, tradition and religion’.
—Jonathan Spall (Financial Times 2010c)
‘[The gold price represents] the longest lasting bubble in human history . . .
[I would not invest my wealth] into something without intrinsic value,
something whose positive value is based on nothing more than a set of selfconfirming beliefs’.
—Willem Buiter (Economist, 2010)
It is generally accepted that there is no direct link between the price of gold and
the balance of supply and demand for the physical commodity. This is because gold is
82
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
almost indestructible and there is a large inventory of the metal held above ground. This
means that a sudden increase in demand could easily be met out of existing supplies. For
example, a net increase in physical supply over demand could still be associated with
an increase in price. If there was, say, an increase in speculative investment demand as
well a weakening in the dollar, the metal would become relatively cheaper for non-USD
investors.
‘It is hard to put an objective valuation on gold, as demand for it is essentially
irrational – its value is in the eye of the beholder’.
—John Authers
4.2.1
The price of gold
Like most metals gold is denominated in US dollars (USD) and in volume terms it is
traded in troy ounces. A troy ounce is slightly less than a conventional avoirdupois
ounce. The maximum nominal (i.e. non-inflation adjusted) price of gold occurred on 5
August 2020 when it reached a level of USD 2,048.15.
4.2.2
Supply of gold
Production of gold over the past 6,000 years has totaled around 187,0003 tonnes of which
approximately 80% is still in existence with the remaining balance probably lost at sea.
However, most of the gold has been produced in relatively recent times with 90% of all
gold having been mined since the Californian gold rush of the mid-nineteenth century.
After central bank attempts to control the market value of the metal proved to be futile,
true liberalisation of the gold price really started to be seen in the 1960s and 1970s. This
prompted a massive boom in production of the metal and is evidenced by the fact that
60% of all gold has been mined since the 1950s. At this time, the metal was trading
at about USD 35.00 per troy ounce and by 2020 had reached its all-time peak of USD
2,048.15.
In the late 1990s as the price of the metal fell back many producers reduced the
amount of expenditure on exploration and so production in subsequent years was relatively flat. It is also noticeable that the production of gold has experienced a certain
geographical drift away from the mature markets of South Africa and North America to
new areas of production such as Asia, Latin America, and Russia. It is also worth noting that reserves of the metal are becoming more difficult and expensive to extract. For
example, in South Africa, the producers are already operating at depths of 4 kilometres.
South Africa has traditionally been thought of as the largest producer of the metal.
However, over the last 20 years they have seen production fall in both absolute and
percentage terms. In the early 1970s the country was producing about 1,000 tonnes per
annum, which accounted for about 70% of all world production. By 2019 their output
had fallen to 118 tonnes, which represented about 3% of total mine supply. The largest
3
Source: World Gold Council (www.gold.org)
83
Gold
supplier of gold is now China who produced about 383 tonnes of metal in 20194 (about
11% of total world production).
Figure 4.2 shows the distribution of global production.
How much gold is left? In the first edition of this text, a figure produced by the
World Gold Council (whose membership includes gold mining companies) estimated
that in 2002, there were 1,227 million ounces or just over 38,000 tonnes of the metal
remaining in the ground. This represented about 15 years of production. A figure
produced in 2020 by the US Geological Society (USGS) suggested that the figure was
currently 53,000 tonnes, which is about 14 years of current production. This would
suggest that there is still a relative abundance of gold, albeit at significant depths
beneath the earth’s surface.
Figure 4.3 shows the way in which supply has evolved over the period from 1980 to
2016. The key features are:
▪ Overall production from mines has continued to rise and is still the major source
of new supply.
▪ After many years, official sales by central banks are no longer a key component of
the available supply; this sector of the market has become a source of demand.
▪ As the price of the metal has risen there has been more incentive to recycle gold and
hence the relative importance of scrap has increased. There are, however, regional
South Africa
118
China
383.2
Ghana
142
Peru
143
Canada
182.9
Russia
329.5
USA
200.2
Australia
325.1
FIGURE 4.2 Production of gold by geographic location
(tonnes).
Source: World Gold Council
4
Source: World Gold Council (www.gold.org)
84
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
differences. For example, in Asia it is quite common to trade in old or unfashionable
pieces for new items unlike the West, where bid-offer spreads for this activity are
not as attractive.
▪ Producer hedging activities were a source of supply until 1999. Thereafter this sector underwent a period of dehedging apart from the period 2014 to 2016.
It is not automatically obvious why the act of selling their production forward by
mining companies contributes to an increase in supply5 . On the other side of the transaction there will be an intermediary such as a trading house or investment bank, who
are long gold for forward delivery. If this exposure is unhedged, they will lose money
if prices were to fall. To hedge this exposure, this intermediary will borrow gold in the
leasing market and sell for spot value. When the forward matures the producer delivers
the gold at a fixed price to the intermediary who uses this metal to repay their borrowing. It is this borrowing by the intermediary that introduces more gold onto the spot
market therefore increasing its supply. This could of course work in the opposite direction. As the price of gold started to rise in the late 1990s and early twenty-first century,
mining companies that had fixed the price of the metal for forward delivery at relatively
low prices came under pressure from shareholders to unwind these transactions. As the
positions were bought back, the opposite effect occurred; the intermediaries bought the
metal for spot value and then lent the gold. The combination of this spot purchase and
subsequent leasing activity resulted not only in an increase in demand but also led to an
6,000
Producer hedging
5,000
4,000
Recycled
3,000
Official sales
2,000
Mine production
1,000
0
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Mine Production
Official Sales
Recycled
FIGURE 4.3 Components of gold supply 1980 – 2019 (tonnes).
Source: World Gold Council
5
A detailed example is given in section 4.4.1.1.
Producer hedging
85
Gold
increase in the price of the metal creating something of a virtuous circle for producers.
Another consequence was that it pushed lease rates to all time historic lows.
4.2.3
The Demand for gold
‘Gold is a need of the people. It is not a luxury item; it is an essential’.
—Mumbai Gold Merchant (Financial Times, 2016)
The factors that influence the demand for gold include:
▪ Investor activity
▪ Industrial uses
▪ Official central bank purchases
▪ Jewellery
▪ Producer hedging activities
Since 1980, the use of gold for investment purposes has been very volatile. However,
post the financial crisis, which coincided with a sharp increase in the gold price, investor
interest in gold increased substantially. One of the major sources of investment demand
has been from exchange traded products (ETPs), which will be considered in Chapter
14. The largest component of demand is traditionally jewellery, but it is very sensitive
to increases in the price of gold.
One of the points shown on Figure 4.4 is the changing role of central banks in the
demand and supply equation. Like most financial institutions central banks will have
6,000
5,000
Producer dehedging
Investment
4,000
3,000
Fabrication
Official
purchases
2,000
Jewellery
1,000
0
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Jewellery
Official Purchases
Other Fabrication
FIGURE 4.4 Components of gold demand 1980–2019.
Source: World Gold Council
Investment
Producer Dehedging
86
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
a portfolio of assets, which have traditionally included gold. The argument for holding gold is that it is not affected by the economic policies of any one country; holding
the asset acts not only as a valuable diversifier but also gives an element of economic
security. It has also been an effective hedge against inflation and currency weakness.
However, some governments have questioned their need to hold gold since it does not
intrinsically earn any income. Although there is an active market for leasing gold, the
returns from this may be below alternative investments.
As of October 2020, the Federal Reserve in the USA still has the largest holding
of reserves at some 8,134 tonnes, with Germany holding the second largest amount
(3,362 tonnes) (see Figure 4.5). The United Kingdom has been one country that has
been seeking to manage its gold reserves more actively and now only ranks 18th overall
with a holding of about 310 tonnes6 . One of the main features of central bank holdings
over recent times has been the increase in holdings by emerging market participants.
For example, China has increased its holdings from 600 tonnes in early 2009 to
1,948 tonnes by mid-2020.
During the 1980s governments purchased a net amount of over 600 tonnes, but in
the following two decades they disposed of about 4,400 tonnes on a net basis figure
(4.6). Following the sharp rise in gold prices in the mid ‘noughties’ the trend changed,
and they became net purchasers of gold. Figure 4.6 shows the reported central bank
activity over the period from 1980 to 2019 showing the annual purchase and sales, and
the cumulative activity.
With the central banks holding a significant proportion of metal above ground their
selling activity in the 1980s and 1990s pushed the price of gold down towards USD
250 per ounce. As a result, pressure from producers led to the first central banks gold
France
2,436
Italy
2,452
USA
8,134
IMF
2,814
Germany
3,362
FIGURE 4.5 Central banks gold holdings (tonnes;
figures as of October 2020).
Source: World Gold Council
6
A detailed example is given in section 4.4.1.1.
87
Gold
1,000
800
0
600
-1,000
400
-2,000
200
-3,000
0
-4,000
-200
-5,000
-400
-6,000
-600
-800
-7,000
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Official sector sales (RHS)
Cumulative activity (LHS)
FIGURE 4.6 Central banks sales / purchases 1980–2019.
Source: World Gold Council
agreement (‘The Washington Agreement’) in 1999. The original agreement existed for
five years and was signed by 15 central banks. The agreement was designed to avoid
excessive fluctuations in the price of gold and has been extended three further times.
The latest agreement was signed in 2014 by the European Central Bank and 20 other
central banks and will last until 2019. The main features of the agreement are:
▪ Gold will still be used as part of the signatories’ monetary reserves.
▪ The signatories will continue to coordinate their gold transactions to avoid market
disruptions.
▪ The signatories to the agreement currently do not have any plans to sell significant
amounts of gold. The annual limit for sales has been set at 400 tonnes for the third
and fourth agreement.
Jewellery demand has risen over the period illustrated in Figure 4.4 but peak
demand was reached in 1997 (3,350 tonnes). The current figure (end of year 2019) is
2,137 tonnes, which accounts for about 44% of all demand. The general increase in
demand until the late 1990s was largely attributable to the long-term decline in the
gold price from a peak value in 1980. However, from then on, the demand for jewellery
has declined as the price of gold has risen from relatively low values. As the price of
gold increased there was a move away from jewellery to gold for investment purposes.
In 1980, gold experienced a significant nominal peak largely because of the Russian
invasion of Afghanistan. At this time investment demand accounted for over 360 tonnes,
88
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
which was about 70% of total demand. During the 1980s it averaged 30% but during the
1990s it only accounted for only about 8% of the total. By 2001, the figure had fallen to
6%, but with the rise in the price of gold during the next decade it increased to reach
a peak of 1,763 tonnes by 2011 (39% of demand). Part of the reason for the increase
in investment demand has been the introduction of exchange traded products (ETPs).
These are securities that track the price of gold and can be traded on a stock exchange.7
4.2.4
Gold price relationships
Since gold can be used to produce tangible goods, the price of gold will be affected by
fundamental demand and supply dynamics but because it also traded as a currency then
to an extent it will also be influenced by macroeconomic and financial drivers.
‘Gold does not protect against inflation or deflation per se but it is likely to
do well in either scenario as it is a hedge against financial dislocation and
uncertainty’.
—Jonathan Spall (Financial Times 2010c)
Inflation
One of the classic theories is that gold is a good hedge for inflation. Before considering
the validity of this statement it is necessary to digress to consider a few aspects of the
traded inflation market. One of the most popular relationships used in the inflation
markets is the Fisher equation. In a simplified shorthand form it states:
Nominal yields = Real yield + expected inflation + risk premium
To explain each of these factors consider the following simple example. Suppose
an investor has USD 100.00 to invest for a 12-month period. Her bank offers her
a deposit rate of 3%; this is the nominal yield. According to the Fisher equation,
this nominal yield can be decomposed into three individual components. The first
is expected inflation which if one assumes to be, say, 2% p.a. then what costs USD
100.00 today will cost USD 102.00 in 12 months’ time. The second component is
the risk premium, which is an additional return that an investor demands for unexpected inflation; for ease of illustration this will be assumed to be zero. In practice
the inflation market tends to combine inflation expectations with the risk premium
referring to the measure as ‘breakeven inflation’8 . The final component is the real yield,
which based on the previous figures returns a residual value of 1%. But defining the
real yield as a residual element does not convey any meaningful sense to this element
of the equation. A real yield signals how much today’s savings are worth in terms of
7
See Chapter 14.
In reality nominal yields rates are impacted by several factors so it is more accurate to describe
the difference between nominal and real yields as the breakeven spread.
8
89
Gold
additional future consumption. So, a basket of goods and services costing USD 100.00
at the start of the period would cost USD 102.00 one year later. By forgoing immediate
consumption and saving USD 100.00 means the investor ends up with a cash flow of
USD 103.00 at the end of the period and so can afford approximately 1%9 more goods
and services.
One aspect of recent times is the existence of negative real yields. Using the same
example, suppose that nominal rates are 1%, inflation expectations are 3%, and the risk
premium remains at 0%. This suggests that real rates are −2%. The investor saves USD
100.00 for 12 months and receives a cash flow of USD 101.00. However, the basket of
goods and services now costs USD 103.00. This means the investor is now worse off by
saving. Logically, they should have spent the money at the start of the period; saving for
the proverbial ‘rainy day’ is not a wise move in this case!
Returning to gold’s relationship with inflation, the question that needs to be asked
is what type of inflation; historic or expected? Presumably when investors buy gold as an
inflationary hedge, most hold some expectation of how inflation will evolve. Therefore,
comparing the evolution of gold prices with historic inflation values may not be valid.
In Figure 4.8 the price of gold is plotted against inflation breakevens10 (i.e. expected
inflation).
The chart breaks down the relationship between the two variables on a year-by-year
basis. The relationship is not clear-cut. Arguably at low levels of expected inflation there
is no positive relationship between the two variables (2014–2017). Only in 2013 is the
relationship relatively pronounced and casual empiricism suggests that in 2011 there
appears to be no correlation.
2000.000
1900.000
1800.000
1700.000
1600.000
1500.000
1400.000
1300.000
1200.000
1100.000
1000.000
1.500
1.700
1.900
2.100
2017
2.300
2016
2015
2.500
2.700
2014
2013
2.900
2012
3.100
3.300
3.500
2011
FIGURE 4.7 Gold price plotted against inflation breakevens.
Source: Data sourced from Barclays Bank
9
Strictly speaking it is 0.98% (USD 103.00 / 102.00 −1 × 100).
The breakeven rate used is the five-year zero coupon inflation swap rate, five years forward.
This measure is used by central banks such as the Federal Reserve to monitor how inflation
expectations are evolving.
10
90
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
2,000
1,900
1,800
1,700
1,600
1,500
1,400
1,300
1,200
1,100
1,000
-2.00
-1.50
-1.00
-0.50
0.00
0.50
1.00
FIGURE 4.8 Gold price plotted against real interest rates.
Source: Barclays Bank
Figure 4.8 shows a more interesting relationship and that is a scatter graph of gold
against real interest rates.11 Here there is much clearer relationship; as real rates fall,
the price of gold will tend to rise. A possible explanation is that since gold does not
pay dividends or coupons (although it is possible to lend gold and earn a relatively low
nominal lease rate) a rise in real rates would encourage investors to switch out of gold
into higher real yield earning assets.
US Dollar
Another popular relationship is the price of gold against the US dollar. Indeed, many
participants treat the USDXAU (US dollar vs. Gold) relationship as a currency pair as
they would say USDJPY. Gold’s role as a currency relates back to earlier days when
coins were made from gold and then attempts by governments to link monetary policy to the metal. In addition, there is the observed investor behaviour that in times of
economic turmoil investors fear that other debts may not be repaid and so will seek out
‘safe havens’ such as gold to offset other potential losses.
The turnover in gold is quite difficult to compare to currency markets. It is a
much smaller market than foreign exchange and a former colleague of the author once
described gold as being like CAD or CHF in terms of volumes traded.
Modeling the gold price
An interesting approach to modeling the gold price was published by Barclays (2013).
They argue:
‘compared with other metals gold is a particularly complex commodity to
model. Not only is it affected by demand and supply dynamics, but it also plays
the role of a currency and, as a result, is influenced by several macroeconomic
and financial drivers. Furthermore, gold is sentiment driven, making it
11
The measure of real rates is five-year real rate swap rates.
91
Gold
important to assess investor confidence. Large above ground inventories
(given that all the gold that has ever been mined still exists in some form)
weaken the link between gold prices and the supply and demand balance. In
the short run, gold moves away from the market balance equilibrium, mainly
because of its currency role. This is inevitably linked to economic indicators,
the state of financial markets, market confidence and political tensions’.
Their model included the following factors as explanatory variables for the spot
price of gold:
Highly significant explanatory variables
▪ Price momentum (positive relationship).
▪ US dollar index (negative relationship).
▪ Physically backed gold exchange traded products (ETP) holdings (positive relationship).
▪ Euro Stoxx 50 index (negative relationship).
▪ MSCI emerging market index (positive relationship).
Significant explanatory variables
▪ US 10-year breakevens (positive relationship).
▪ University of Michigan Consumer sentiment index (negative relationship).
▪ US 10-year real yields (negative relationship).
▪ US presidential approval ratings (negative relationship).
▪ An in-house measure of industrial production from 30 countries (positive relationship).
4.3
THE GOLD LEASING AND DEPOSIT MARKET
One of the factors that distinguish the gold market from that of, say, base metals, is the
large amount of metal that already exists above ground. This excess of metal has led to
the development of an active leasing market12 . In a leasing transaction, the lender will
temporarily transfer title of the gold to another entity for an agreed period in return
for earning interest either in gold or in cash. From the lender’s point of view this could
be a useful way of earning income on an asset that otherwise has no intrinsic return.
However, it should always be borne in mind that an opportunity cost exists as gold interest rates are generally lower than unsecured interbank rates. The holder of gold can
always sell the metal and invest the proceeds in a simple USD deposit. And with pressure coming to bear on central banks to make gold ‘earn its keep’, many entities have
been tempted to go down this path, particularly in a low lease rate environment.
Although the main purpose of the leasing market is to borrow and lend metal,
care must be taken when using these terms within other metals markets. For example,
12
Some market participants may use the term ‘deposit’ in relation to this market.
92
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
within the base metals market, ‘a borrow’ is a simultaneous purchase and sale for different value dates. This type of transaction will be analysed in greater detail in the base
metals chapter. Therefore, the terms ‘leasing’ and ‘deposits’ should be used with respect
to the gold market to avoid confusion.
To understand the subtleties of the leasing market, it is instructive to take a sidestep
and consider how gold forward prices are formed.
4.3.1
Forward price formation
To illustrate the principles of forward price formation, take the example of a gold producer who wishes to achieve a degree of price certainty over future production13 .
The producer approaches a bank asking for a price for delivery of gold in, say, six
months. There is an old derivatives adage that states that the price of any product will
be driven by the cost of hedging the bank’s own exposure – ‘if you can hedge it, you can
price it!’. If the bank does not hedge itself then in six months’ time it will take delivery
of gold at the pre-agreed price and will then be holding an asset whose current market
value could be lower (or higher) than the price paid to the producer.
To avoid the risk that the price of gold will fall, the bank executes a series of transactions on the trade date that will mitigate this risk. Since the bank is agreeing to receive
a fixed amount of gold in the future, it sells the same amount in the spot market. However, the bank has sold a quantity of metal now that it will not take delivery of until a
future period; six months in this case. To fulfill the spot commitment, it leases the gold
until the maturity of the contract with the producer.
Having sold the gold spot and borrowed to cover the sale, the bank is now left holding dollars. Since the bank would be looking to manage its cash balances efficiently,
these dollars would now be invested until the producer delivers the gold in six months’
time. As a result, it is possible at the inception of the trade to identify all the associated
hedging cash flows, allowing the bank to quote a price that will ensure no loss at the
point of delivery. The maturing principal plus interest on the USD cash deposit is used
to pay the producer on the maturity of the contract. The gold received from the producer
is used to repay the gold leased from the central bank. The maximum amount the bank
will pay the producer cannot exceed the proceeds received from the initial spot sale of
the metal plus the interest received from the dollar deposit less the fee to the lender
of gold.
A numerical example may help illustrate the point. We will assume that the producer asks for a six-month (182 days) forward price. For simplicity we will base the
calculations on a single ounce and use a spot price of USD 1,300. The trader sells the
metal for spot value at USD 1,300 and to complete the delivery, he borrows from the local
central bank for six months at a lease rate of 0.25% per annum. The dollars received from
the spot sale are put on deposit for six months at LIBOR to earn the prevailing rate of
2% per annum. Performing a quick calculation the trader works out that the borrowing
13
It may well be that the hedging of future revenues may also be associated with some form of
bank financing. That is, the bank will only lend money to a producer if it takes steps to ensure
the stability of future income to service the debt.
93
Gold
fee equates to USD 1.64 (spot price x lease rate x 182/360) and that he will earn USD
13.14 from the deposit (spot sale proceeds x LIBOR x 182/360). The maximum amount
the bank can afford to pay the producer is USD 1,311.50. This is calculated as spot sale
proceeds plus interest on the LIBOR deposit minus the leasing cost.
The forward price is simply the spot price plus the cost of carrying an underlying
hedge. It is important to note that the shorter the time to maturity the smaller the differential will be between the spot and forward price since the hedge is carried for a
shorter period. Indeed, if every day we were to recalculate the forward price applicable
for a single fixed date in the future, the differential would reduce (all other things being
equal). By the spot value date the two prices will have converged, although, in reality
this convergence would not happen in a linear fashion as it is dependent on how the
other components of the forward price move (e.g. LIBOR and the lease rate).
Note that in this example, the forward price is greater than the spot price. This does
not means that the bank thinks that the gold price will increase over time or that it is the
market’s best guess of future spot prices. The forward price is a mathematical construct
that reflects the income and expense from hedging the underlying exposure.
In the interbank market the convention for quoting forward prices is similar to that
seen in the FX markets – i.e. interbank participants do not quote the forward price (e.g.
USD 1,311.50), but quote the difference between spot and forward prices as a percentage
per annum. For this example, the quote would be:
= (Forward price∕spot price) − 1 × 360∕182
= (1, 311.50∕1, 300) − 1 × 360∕182
= 1.7498%
This differential has a special name in the gold market and is variously called the
GOFO rate/swap rate/contango bid. GOFO is a shortened version of Gold Forward
Offered Rate.
By way of example the following hypothetical quotes shown in Table 4.2 are based
on a mid-market spot rate of USD 1,331.
TABLE 4.2 Forward gold prices.
Maturity
Swap bid
Swap offer
Outright bid
Outright offer
1 week
2 weeks
1 month
2 month
3 month
6 month
9 month
12 month
24 months
1.6000
1.6250
1.6600
1.7300
1.7900
1.8700
1.9700
2.0700
2.2350
1.8000
1.7750
1.7600
1.8300
1.8900
1.9700
2.0700
2.1700
2.3850
1331.06
1331.49
1332.55
1.334.42
1336.67
1343.23
1350.53
1358.58
1390.96
1331.30
1331.75
1332.85
1334.82
1337.19
1344.08
1351.72
1360.11
1395.19
94
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Note that term structure of outright prices rises with respect to maturity. This is a
reflection that, in all cases, LIBOR exceeded lease rates and so the swap rate (‘GOFO’)
is positive. It is rare for swap rates to be negative; this would only happen if lease rates
were greater than LIBOR.
4.3.2
Deriving implied lease rates
When deriving the forward price for gold it was assumed that spot, LIBOR, and the lease
rate were all observable. However, data on actual leased transactions is not routinely
reported and so the lease rate can be inferred from observable values. Since:
Forward price = spot price + LIBOR − Lease rate
(4.1)
Forward price − spot price = LIBOR − Lease rate
(4.2)
Rearranging:
Since the difference between spot and forward prices is the swap rate, then
equation (4.2) can be rewritten as:
Swap rate = LIBOR − lease rate
(4.3)
Swap rate − LIBOR = lease rate
(4.4)
So,
If the six-month gold swap rate were 2% p.a. and six-month LIBOR was 1.5% p.a.,
it would imply a lease rate of 0.50% p.a.
4.3.3
Who lends and borrows gold?
On the supply side of the lease market the main players include central banks and institutional investors, both of whom are looking to enhance the return on their physical
holdings. On the demand side a variety of factors could impact the level of lease rates.
They include:
▪ The level of money market interest rates (e.g. LIBOR, although this relationship
has experienced times of dislocation).
▪ Producer hedging programmes.
▪ Demand from jewellery manufacturers.
▪ Demand from speculative short selling.
Firstly, there is the absolute level of interest rates. It would seem reasonable to
assume that a fall in interest rates will cause lease rates to fall. Secondly, one of the key
reasons why gold is borrowed is to facilitate producer hedging programmes. As producers reduce the hedging of their mined product, the lease rates begin to fall. This
was covered in the section that analysed demand and supply price drivers. The lease
95
Gold
market displays a term structure with different rates being charged depending on the
maturity of the transaction. As a rule of thumb, producer hedging activity tends to drive
the longer end of the curve towards the 12-month maturity and beyond, while central
bank lending activity tends to be focused on shorter maturities. A change in the level of
activity of either of these entities can cause the shape of the curve to change.
However, there are exceptions to this generalised explanation of how the curve
moves. For example, certain producer hedging products (e.g. a floating rate forward)
derive their value from shorter-term rates and therefore drive that part of the curve.
Another factor influencing the level of lease rates is the utilisation of the market by
refiners or fabricators. Take a jewellery fabricator; in normal circumstances one would
expect them to have to buy the gold and then transform it into jewellery. However, they
may have to pay for the raw materials before selling the gold. This may involve having
to borrow US dollars and will leave them open to an element of price risk (i.e. the difference between what they pay for the raw material and what they eventually raise from
the final sale).
An alternative strategy is for the fabricator to borrow gold from a bank and use that
to produce the final product. When the product is sold, they can then use the proceeds
of the sale to buy the physical gold from the spot market and repay the gold loan. In this
way the price risk is eliminated and the cost of leasing the gold is lower than the US
dollar LIBOR rate they will have paid to finance a purchase of metal from a producer.
If a speculator believed that the price of gold was going to fall then he could sell the
metal for spot delivery, borrow it on the leasing market to fulfill the sale, and wait for
the price of gold to fall before buying it back from the market, hopefully at a lower price.
The purchase is used to repay the loan of the metal. The profit to the short seller would
be the difference between their buy and sell price less the lease rate paid. However, in
a rising price scenario this strategy may be less popular, so no demand for borrowing
gold is created, putting further downward pressure on lease rates.
It is also possible for the lease payments to be made in gold, so called ‘gold in gold’
contracts. The interest rates for these transactions differ slightly from those charged
for cash transactions. Assume we have a one-year lease rate (zero coupon, actual/360
day basis) of 0.75%, a gold swap rate of 3.75% (defined here as the annual percentage
difference between the spot and the forward rate), and a spot gold price of USD 1,300.
Let us assume that a financial institution lends out 100,000 ounces for this one-year
period (365 days). This would give the following payment in ounces (rounded):
100,000 × 0.75% × 365∕360 = 760 oz.
The monetary equivalent for forward value in one year’s time can be determined by
applying today’s one-year forward price to the number of ounces settled. The one-year
forward price is calculated as:
Spot price × [1 + (swap rate × days in period∕360)]
USD 1,300 × [1 + (0.0375 × 365∕360)] = USD 1,349.43
Which, when applied to the settlement amount in ounces, gives a forward value of
USD 1,025,566.80 (760 ounces x USD 1,349.43).
96
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
If the same transaction were done for payment in cash, the amount to be repaid at
the contract’s maturity would be based on the current spot price:
100,000 × USD 1,300 × 0.75% × 365∕360 = USD 988,541.67
To make both transactions equal in monetary terms one year forward, this would
mean that the lease rate for ‘gold in gold’ transactions would have to be 0.722527%. This
is calculated as:
[(USD 988, 541.67∕USD 1, 349.43) × (360∕365)]∕100,000
4.4
HEDGING
4.4.1
Forwards
Arguably, one of the main risk management issues for producers is the amount of revenue generated from either current or future production. From a logistical perspective
the producer needs to match the timing of hedging transactions with planned future
mining activity. The interested reader is referred to the research document written by
Jessica Cross (2001) for an outline of producer motivations and hedging philosophies.
Producers have typically favoured hedging products that are easy to understand, represent ‘good value’, can be processed using their existing computer systems, and can
attract favourable accounting treatment. It comes as no surprise to find that simple
‘vanilla’ forward and option products are often the most preferred structures.
It is worth noting at this point that forwards could come in different variations. It is
also possible to consider two other possibilities:
1. The exchange of a pre-agreed amount of cash for a pre-agreed amount of gold at
a pre-agreed date where delivery will require the actual physical movement of the
metal. For example, a producer sells forward new mine production in exchange for
a fixed USD amount.
2. The exchange of a pre-agreed amount of cash for a pre-agreed amount of gold at
a pre-agreed date, where delivery is done by ‘book entry’. For example, two banks
trade unallocated gold and so the metal stays in a vault but the client’s respective
holdings are updated accordingly.
Investment banks have developed a wide range of variations based on the humble
forward. A comprehensive analysis of the relative merits of each product is provided
in the publication ‘Gold Derivatives: the market view’ (Cross, 2001). The following is a
short description of some of the main products. With all these deal types the structuring
bank simply alters the constituents of the forward price (LIBOR, lease rate, maturity),
which allows the producer more flexibility to express views on the direction of certain
market parameters.
97
Gold
Spot deferred – in this transaction the two parties agree the spot price at the
outset. However, there is no fixed maturity, and the deal can be terminated with the
appropriate amount of notice. The final forward price is based on the daily movement
of the LIBOR and lease rates. This is in effect a rolling contract where gold is initially
sold for spot value (trade date plus two business days). The following day, the position is
rolled forward by buying back and then reselling for spot value. This effectively pushes
the settlement date one day further forward. As part of the transaction the dollar
proceeds earned from the sale are invested on an overnight basis while lease costs are
incurred to cover the short sale. The transaction settles daily with a debit/credit being
made to the customer’s account based on the movement of LIBOR and the lease rates.
This structure allows the producer to earn the short-term contango.
Floating rate forwards – here the contract fixes the appropriate spot gold price and
LIBOR components of the contango for the maturity of the contract. However, the final
forward price is not calculated until maturity and is based on an average value of a
short-term lease rate over the life of the deal (e.g. the three-month lease rate).
The flat rate forward – in this contract the producer sells its production forward in
stages over an agreed period. For example, the producer may agree to a contract with
a 12-month final maturity with the gold being delivered every three months. Under
normal forward pricing theory this would involve a different price for each settlement.
However, as the product name suggests the contract is structured in such a way that
each of the four deliveries takes place at a single forward price. The single price is based
on the weighted average of the forward prices at each settlement date. The calculation is
the sum of the forward rates at each of the delivery dates weighted by a discount factor
of the same maturity, all divided by the sum of the discount factors.
An example of a structured forward is:
Convertible forward – this can be constructed by combining the purchase of a vanilla
put option and the sale of a reverse knock in call option, both with the same strike. This
transaction is structured to have zero cost by manipulating the barrier on the reverse
knock in call. The strategy is based on the principle of put-call parity to create a synthetic forward. The mathematical form of put-call parity varies according to the type of
underlying asset (see Tompkins, 1994), but can be expressed in a convenient way as:
+C − P = +F
That is, the purchase of a call (+C) and the sale of a put (−P) with the same strike,
notional and maturity will be equivalent to having purchased the underlying asset (+F).
For the convertible forward structure, the initial position is that the producer owns a
given amount of the physical gold in addition to which they have bought protection
against a price decline by buying a put (+P). In addition, the producer sells a reverse
knock in call option (−C). This barrier option has a trigger point placed above the spot
rate, which will activate the option if the price of gold rises. This makes the barrier an
‘up and in’ call option. If the trigger is hit activating the short call, combining this with
a long put will give the producer a synthetic short forward position in gold (e.g. −C+P
= −F), which when combined with their long physical holding in the metal will ensure
them of a fixed delivery price.
98
4.4.1.1
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Unwinding forward sales
The value of any forward position will change as the underlying price drivers change.
For example, a producer may be tempted to unwind a forward hedge if the spot price
of gold has risen substantially, since it would now be better to sell the gold in the spot
market. Indeed, several gold mine producers (e.g. Barrick in 2009) faced shareholder
pressure to unwind long-dated forward hedges as the price of gold rose. Given that many
banks will now require collateral to be deposited for profitable transactions a rise in the
gold price will have required producers to deposit increasing amounts of cash which
will have caused a strain on their cash flow.
Say the producer enters a three year forward at a pre-agreed fixed price for the sale
of gold at a price of USD 500.00/oz. Assume that two years later, the price of gold has
risen substantially such that the prevailing one year forward price is USD 1,100. The
existing hedge position is now unfavourable relative to the underlying market. The producer contacts an investment bank requesting a price to unwind the hedge. This ‘break
cost’ will simply be the ‘mark to market’ (i.e. the current value) of the transaction. The
producer has an agreement to deliver a fixed amount of gold for a fixed price on an
agreed fixed date. To neutralise this exposure they could do a reversing deal, where for
the same maturity date the producer could agree to buy the same amount of gold and
pay a fixed sum of money. The break cost is simply the difference in value between
the original deal and the reversing position, which is settled in cash. This break cost is
sometimes referred to as an ‘exit price’14 .
Using the figures quoted above, the producer would incur a gross termination cost
of USD 600.00 (USD 500.00 − USD 1,100). Since this maturity is still one year away this
would need to be present valued using an appropriate one-year zero coupon rate. If it
is assumed that the appropriate rate is 3% and there is exactly one year to maturity, the
settlement amount is:
USD 600.00∕(1.03) = USD 582.52
This represents a mark-to-market loss from the producer’s perspective. If the producer were a non-USD entity this would then have to be translated back to their domestic currency, using the one-year forward currency rate.
4.4.1.2
Gold swaps to alter a delivery date
Suppose that a gold producer has a legacy outright forward sale in effect that requires
them to deliver 50,000 ounces on an agreed fixed date at a price of USD 1,300/oz. The
producer realises that due to operational issues they will be unable to deliver the gold on
the required date and so contacts the contracting bank to defer the delivery date. This
deferral can be achieved using a gold swap traded two days before the original forward
14
The exit price is usually defined as the price that would be received to sell an asset or paid
to transfer a liability in an orderly transaction between market participants at the measurement
date.
99
Gold
deal becomes due. The following example is done from the client’s perspective and the
cash flows are based on the values in Table 4.2.
Original outright forward (due to settle in two days’ time)
Gold leg
Sell 50,000 oz.
USD leg
Receive USD 1,300/oz.
Gold swap
Spot
Six month
Buy 50,000 oz.
Sell 50,000 oz.
Pay USD 1,330/oz.
Receive USD 1,343/oz.
The spot leg of the swap is designed to reverse the original outright forward transaction. On this date, the amount of gold to be dealt nets off and the customer will be
required to pay USD 30.00/oz. to the bank to reflect the movement in the gold price.
The forward leg of the swap reestablishes the economics of the original transaction in
that the customer now has a commitment to deliver the gold in six months’ time at a
fixed price of USD 1,343/oz.
4.4.2
Swaps
The use of word ‘swap’ without any qualification in the gold market should be treated
with caution as it could have several meanings. Conventional commodity swaps are
possible in the gold market and these comprise of cash-settled, Asian-style, fixed for
floating structures. A structured swap would typically include some form of embedded optionality that results in a payoff that is initially more favourable than the vanilla
equivalent. However, if the optionality is exercised then the participant is faced with a
payout that is less favourable.
4.4.2.1
Locational and quality swaps
Two types of swaps that are perhaps unique to the gold market are quality and locational
swaps. When reference is made to ‘the’ gold price then it can usually be taken to mean
the ‘loco London’ price. Other loco prices are possible such as Zurich or areas with
common transport and warehousing facilities (e.g. Tokyo, Shanghai) as well as where
the refineries are based. For example, as I type this text, the price of gold loco London
is USD 1,341, loco Zurich USD 1,342, and loco New York USD 1,340.
Suppose that a US mining company wishes to swap 50,000 ounces of its physical
American gold holdings for a similar position in London. It contacts its custodian bank
100
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
and arranges a locational swap. Typically, these are booked as two separate transactions.
From the client’s perspective they are executing the two following transactions:
Sell 50,000 oz. gold at USD 1,340 loco New York
Buy 50,000 oz. gold at USD 1,341 loco London
The client must pay to the bank the difference in USD between the two loco prices,
which in this case will be USD 50,000. On the sale leg, the mining company’s metal
account in New York is debited with the offsetting credit to the New York gold account
of the transacting bank. On the purchase leg, the transacting bank debits its London
metal account for 50,000 oz. and credits the mining company’s London metal account
for the same amount. In this way the mining company that now owns metal in London
does not actually have to arrange for the metal to be shipped.
A quality swap would work in a similar fashion. In the loco swap example, it was
assumed that the qualities of the two metals were equal. A quality swap would be where
one party owns gold of a given purity, but has a commitment to deliver gold of an alternative quality. Rather than have to pay for the metal to be refined into the necessary
grade it may be possible to enter into a swap transaction where ownership of gold is
transferred with a settlement in USD to reflect any price differential.
4.4.2.2
American barrier reset swap
A possible term sheet is presented below:
Fixed price payer:
Floating price payer:
Fixed price:
Floating price:
Reset event:
Trigger level:
Reset fixed price:
Settlement frequency:
Comparable vanilla swap level:
Client
Bank
USD 1,250, subject to a reset event
Unweighted arithmetic mean based on the daily
AM LBMA gold price over each month.
If the LBMA gold price settles at or above the
trigger level, the fixed price for the current
month is set to the reset fixed price.
USD 1,400
USD 1,400
Monthly
USD 1,300
The structure is illustrated in Figure 4.9.
The consumer buys physical gold monthly from their usual supplier and for ease
of illustration assume the commercial contract also references the LBMA gold price.
As is common with many commodities contracts, the agreed price will be based on the
concept of ‘average of the month’ prices. This means that the consumer is protected
101
Gold
Floating price
Client
Bank
Fixed price
(subject to reset)
Floating price
Physical gold
FIGURE 4.9 Structure of American barrier reset.
from a sudden upward spike in prices. By the same token they cannot benefit from a
sharp fall in prices.
The main purpose of the swap is to transform the floating price exposure of
the physical supply contract into a fixed price, but the advantage of this structure is
that the fixed price is initially lower than the vanilla equivalent. If the price of gold
increases beyond the trigger level of USD 1,400 then the fixed price payable by the
client increases to this level, which is now higher than the vanilla equivalent.
The reduction in the fixed price is achieved by the fact that the client has sold a
strip of options. The premium receivable on this option is not paid in cash but is instead
used to subsidise the lower fixed rate. The seller of an option is normally faced with
open-ended losses but note that in this case that if the option is exercised then the client
will end up paying a fixed amount of USD 100.00 more than the vanilla equivalent. This
would suggest that the embedded strip of options is a series of digital call options. The
‘barrier’ is actually the strike and the option is American in style and so can be exercised
at any time.
In this structure the first of the embedded options is a spot starting one-month
option with a strike of USD 1,400. The remaining options have the same characteristics
(e.g. one-month maturity, pre-agreed strike of USD 1,400 and a fixed payout of USD
100.00), except they are all delayed start instruments.
If the optionality is exercised in one month, the higher fixed rate will only apply for
this single period. At the start of the next period, the client will pay the advantageous
fixed rate unless the gold price has increased beyond the threshold.
4.4.2.3
American barrier knock out swap
This swap is also structured to offer an improved fixed price to the customer subject to
a trigger event. We will assume the same context as the previous example. A term sheet
may look as follows:
102
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Fixed price payer:
Floating price payer:
Fixed price:
Floating price:
Client
Bank
USD 1,200, subject to a reset event
Unweighted arithmetic mean based on the daily AM
LBMA gold price over each month.
Knock out event:
If the LBMA gold price settles at or above the trigger
level, the swap knocks out and payments cease.
Trigger level:
USD 1,400
Settlement frequency:
Monthly
Comparable vanilla swap level: USD 1,300
In this structure the client initially pays a fixed price of USD 1,200, which is below
the vanilla equivalent. If the price of gold increases beyond the trigger price, then the
remaining portion of the swap terminates, leaving the consumer unhedged in a higher
price environment.
Since the swap can be terminated at any time that the gold price hits the barrier
then it is an American-style structure.
Once again, the client is initially enjoying a subsidised fixed price, which suggests
they are selling some form of optionality. Since the swap will terminate if the option
is activated, this means the client has sold an option on a swap, i.e. a swaption. This
option will require the client to enter a second swap with terms equal but opposite to
those of the initial swap. Suppose that the gold price hits USD 1,400 at the end of the
first month triggering the embedded option. The client will then be required to receive
fixed/pay floating on an offsetting swap with 11 months remaining, but with the same
terms as the initial swap. This means that both the fixed and floating cash flows will
cancel out. However, rather than require the client to hold two positions that are perfectly offsetting, the bank will simply agree to terminate the structure leaving the client
unhedged.
4.4.3
Options
In addition to using forward based strategies, gold producers can hedge their production
using options.
4.4.3.1
Vanilla put options
The simplest structure for a producer is to buy a put option. This gives them the right
but not the obligation to sell a pre-agreed amount of gold to the bank at some future date
at a price agreed today. Let us assume that the producer decides to buy a six-month put
at a strike price of USD 1,300 per ounce with the current forward market at USD 1,306.
This option is out-of-the-money as the strike price is less favourable than the underlying
forward market. Since the producer has bought a European-style option that can only
be exercised upon maturity, the underlying is the forward rather than the spot market.
103
Gold
The put option will give the producer protection if the market falls below this
strike level, but will allow them to walk away from the contract if the price of gold rises.
A common misconception concerning this strategy is that it would be appropriate
if the producer expected the price of gold to fall. If the producer held this view, it
would be cheaper for them to sell the gold forward particularly if the market were
in contango. The purchase of a put would be appropriate if the producer felt the
price of gold were to rise, but wanted insurance in case their view of the market
was wrong.
Although the purchase of a put option is a simple strategy, it is often unpopular due to the premium cost. With implied volatilities for gold often close to 20% (for
short-term options) the premium on a substantial production target could be relatively
high. For example, pricing a six-month option with a strike of USD 1,300, a forward
price of USD 1,306, and an implied volatility of 20% would return a theoretical fair
value of USD 70.00 per ounce. This represents about 5% of the current spot price, which
on a position of, say, 50,000 ounces would represent a substantial upfront cash flow
(USD 3.5 million).
4.4.3.2
Asian style put option
One of the popular approaches to the issue of hedging is to offer options with lower
premiums. The first alternative would be to offer the producer an average rate option
(sometimes referred to as ‘Asian’ options). These are options that settle against the average price of the asset during the option’s lifetime. These can provide an effective hedge
for consumers or producers who buy or sell the physical metal on commercial contracts
that reference some form of averaging process. A typical term sheet from a producer’s
perspective may look as follows:
Volume:
Option type:
Strike:
Settlement:
Maturity:
Payoff:
Averaging:
50,000 oz.
Put option
USD 1,300 (ATM spot)
Cash-settled
Six months
At maturity, (i) if XAU average < Strike price client receives Strike –
XAU average (ii) otherwise no payoff will take place.
XAU average is the average of the daily AM LBMA gold price over
the averaging period.
Asian-style options will always be less expensive than the non-averaging equivalent. The volatility input is still referencing the underlying. However in this instance,
it is not the outright price but an average price over some pre-agreed period. As a
result, the volatility of an average price series is lower than that of just the outright
price. Using an implied volatility of, say, 15% rather than the 20% used in the previous
104
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
example and an averaging period equal to the option’s maturity (i.e. six months) the
premium returned is USD 28.59. It is tempting to claim that this is ‘cheaper’ than
the non-averaging equivalent, but this may be misleading use of the terminology.
‘Cheapness’ is more commonly used to denote options that are undervalued relative
to some notion of ‘fair value’. In cash flow terms it will cost less than a European-style
option, but since its payout references the average of a time series, the payout will also
be lower. As the averaging period falls, the value of the option will increase; taken
to the extreme an averaging period of one day will make it indistinguishable from a
European option and will therefore cost the same.
4.4.3.3
Barrier options
A barrier option is a structure where the payout depends on whether the underlying
price crosses or reaches a defined barrier level. Knock-in options start their lives worthless and only become active in the event a predetermined knock in barrier price is
breached. Knock-out options start their lives active and become null and void in the
event a certain knock-out barrier price is breached.
Volume:
Option type:
Strike:
Settlement:
Maturity:
Knock out barrier:
Payoff:
Premium:
50,000 oz.
Knock out put option (‘up and out’)
USD 1,300 (ATM spot)
Cash-settled
Six months
USD 1,350
At maturity, (i) if spot gold has never traded above the
knock-out barrier during the life of the contract, the client
receives MAX (Strike – XAU final, 0) on the settlement date.
(ii) Otherwise no payoff will take place.
XAU final = the LBMA gold price on the expiration of the
contract.
USD 35.03
From the producer’s perspective they are most concerned about the possibility of
falling prices. If prices were to rise through the barrier, they may feel comfortable with
the termination of the option. It will leave them unhedged if the price of the underlying
asset were then to fall. As was suggested in section 1.8.2 the main motivation for using
barrier options is that the premium is less than the European equivalent. Compared to
the vanilla put option, the premium has fallen by 50%. If the producer asked for the
barrier to be moved closer to the current spot price (say USD 1,330) the premium on
the option would fall as the option is more likely to be terminated. This reduces the
probability that the option will be exercised and as such reduces the cost. If the barrier
is moved further away from the spot price, the option will increase in value. There is
less chance of the option being terminated and so the position more closely resembles
that of a vanilla equivalent.
Another alternative would be a ‘down and in’ put option.
105
Gold
Volume:
Option type:
Strike:
Settlement:
Maturity:
Knock in barrier:
Payoff:
Premium:
50,000 oz.
Knock-in put option (‘down and in’)
USD 1,300 (ATM spot)
Cash-settled
Six months
USD 1,250
At maturity, (i) if spot gold has traded below the knock-in barrier during
the life of the contract, the client receives MAX (Strike – XAU final, 0)
on the settlement date. (ii) Otherwise no payoff will take place.
XAU final = the LBMA gold price on the expiration of the contract.
USD 66.70
Here the producer has downside protection, but only once the option is activated
upon spot trading through the USD 1,250 barrier. At that point, the producer has the
right to sell at a price that is USD 50.00 higher. This means that the option is ‘born’ with
USD 50.00 of intrinsic value.
The premium is lower than the vanilla equivalent but is higher than the knock-out
variant.
4.4.3.4
Two-asset barrier options
Consider a scenario where a gold producer also mines copper as a by-product. The
producer may also seek to hedge this additional production and could feasibly use
options in a similar manner to those outlined above. However, the cost of hedging this
secondary production could be managed using a two-asset barrier option. This type
of option links the payoff on the copper component to the price of gold. It may be
reasonable to suggest that the producer would only need protection against adverse
copper price movements when gold prices are low. A hypothetical term sheet for this
type of option may look as follows:
Option type:
Spot price of copper:
Spot price of gold:
Copper strike price:
Gold knock in barrier:
Settlement:
Maturity:
Payoff:
Premium:
Two asset knock-in put option (‘down and in’)
USD 7,100
USD 1,300
USD 7,100 (ATM spot)
USD 1,000
Cash-settled
Six months
At maturity, (i) if spot gold has traded below the knock-in barrier
during the life of the contract, the client receives MAX (Copper
strike price − Cu final, 0) on the settlement date. (ii) Otherwise no
payoff will take place
Cu final = the London Metal Exchange cash price for copper on the
expiration of the contract.
USD 52.83 (Vanilla six-month copper put option USD 562)
106
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 4.3 Relationship between option premium and
correlation for a two-asset barrier option.
Correlation input
+1.0
0.0
−1.0
Option premium (USD)
93.35
52.83
0.01
If the prices of copper and gold are positively correlated then as the price of gold
falls below the USD 1,000 barrier then the right for the producer to sell copper at the
strike will be activated; the positive correlation in prices would make it likely that the
put option would now be ITM.
Compared to a vanilla copper put option this structure represents a major reduction of the premium. This is because the pricing of this type of option includes a component to reflect the correlation between the two underlying prices. All other things
being equal, the relationship between the premium and the price correlation is positive
(Table 4.3).
If the two prices exhibit positive correlation, then a fall in gold prices is likely to
be associated with a fall in copper prices. This increases the likelihood that the option
will expire ITM. Negative correlation of prices returns a very low premium value. This
relationship would suggest that high gold prices are associated with low copper prices.
Therefore, the barrier is less likely to be activated as a result, but the activation barrier
is less likely to be hit, so the option will expire OTM. Conversely, low gold prices would
be associated with high copper prices and so although the option may activate the right
to sell at price lower than the market would be of no value.
4.5
TRADING GOLD
4.5.1
Gold swaps / FX swaps
Although there are several generic trading strategies a gold trader can execute (e.g. trading option volatility), this section focuses on those strategies that are perhaps somewhat
specific to the gold market. Section 4.3.1. illustrated how a fair forward price could be
derived using time value of money principles and introduced the concept of the gold
swap rates. Within the context of forward markets, it is important to make a distinction
between outrights and swaps. An outright forward is a single transaction where a given
quantity of gold is to be delivered at a fixed date beyond spot value for a pre-agreed fixed
price. A producer could use a single outright forward sale of gold as a way of achieving
cash flow certainty.
The rationale for a gold swap is much less intuitive although readers with a background in foreign exchange may be familiar with the instrument; indeed, some market
participants do refer to them as ‘FX swaps’. A gold swap consists of a spot and forward
transaction that are executed simultaneously. The following example shown in Table 4.4
illustrates the principles, based on a two-month transaction expressed from the perspective of a market maker.
107
Gold
TABLE 4.4 Gold swaps.
Spot price
Swap %
Outright price
Trader’s actions (market
maker’s perspective)
Interpretation
Bid
Offer
1,330.65
1.73%
1,334.42
Sell gold for spot value
Buy gold for two-month
forward value
Trader lends gold, borrows
USD
1,330.83
1.83%
1,334.82
Buy gold for spot value
Sell gold for two-month
forward value
Trader borrows gold, lends
USD
Suppose a market maker executes a trade at their bid price. Typically, these trades
may base the spot value of the transaction on the mid-market spot price, but here the
example will use the bid price. For ease of illustration the example is based on a single
ounce.
Gold component
Cash component
Spot date
Forward date
Bank delivers one oz. of
gold
Bank receives 1,330.65
Bank receives bank one oz.
of gold
Bank repays 1,334.42
When looked at vertically (i.e. thinking of the transaction as a combination of a spot
and forward deal), the rationale for the swap is not obvious. Arguably the better way to
view the transaction is to look at it horizontally. Looking at the cash component, the
bank has taken in USD and then repaid it two months later; economically this is nothing more than a USD borrowing. Looking at the gold element of the transaction the
trader has lent gold for the period. Combining the two suggests that the transaction
could be viewed as nothing more than a USD borrowing collateralised with gold. However, it does beg the question as to why the trader did not simply borrow USD on an
uncollateralised basis. Although borrowing on an unsecured basis in the money markets is possible, logic dictates that borrowing on a collateralised basis should be cheaper.
Indeed, for some participants whose credit rating is considered questionable, this may
be the only way in which they can borrow.
Viewing the transaction as a collateralised loan allows us to reinterpret the swap
rates. At the bid price the percentage difference between the incoming dollars and the
amount repaid represents a borrowing cost of 1.73% p.a. At the bid price the trader
is lending gold and borrowing USD at a net cost of 1.73% p.a. At the offer price the
trader is borrowing gold and therefore lending dollars on which they would expect to
earn 1.83% p.a. net.
Although this analysis conveys one rationale for using gold swaps, it is possible for
anomalies to arise. In the section on leasing it was highlighted that the swap points represented the percentage difference between spot and forward prices. The two factors
108
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
that determined this differential were LIBOR rates and lease rates. It was also pointed
out that the market often inferred the lease rates by subtracting the swap rate from the
prevailing LIBOR rate (‘LIBOR minus GOFO’). Equation (4.3) showed that the swap
rate is calculated as LIBOR minus the lease rate. So, if the lease rate is positive, the
collateralised cost of borrowing (i.e. the swap rate) is less than the unsecured equivalent. Using the figures in the above table, suppose the appropriate value for two-month
LIBOR was 1.63% implying a lease rate of −0.10%. This suggests that if you were to
borrow some gold you would be paid to do so!
However, one possible explanation for this can be seen in the FX market and is
referred to as the FX basis. Suppose that if a bank were to borrow in the unsecured
LIBOR market it would be required to pay a margin over the benchmark LIBOR rate
because of credit concerns. The trader decides to use the gold market to raise collateralised funds knowing that her borrowing cost will be lower. If there were sufficient
demand for this type of USD borrowing, then the amount of USD that counterparties
will demand to be repaid in the forward leg will increase. If the observed LIBOR and
the spot rate are assumed to be unchanged, the increase in the forward price will be
associated with a lower implied lease rate. If the demand to borrow USD is sufficiently
high, then eventually the implied lease rate will turn negative.
A simple example will illustrate based on the following market conditions:
▪ Tenor: 12 months.
▪ Spot price: USD 1,300.
▪ 12-month LIBOR: 3%.
▪ 12-month lease rate: 0.10%.
▪ Day basis: Actual/Actual.
▪ Implied 12-month forward price: USD 1,337.70.
Based on the above figures the 12-month net borrowing cost is 2.90%, which will
also be the swap rate. Suppose now that there is an instantaneous increase in demand
to borrow USD using gold as collateral. Traders will react to this demand by increasing
the cash amount that must be repaid at maturity. Assume that the swap rate is now
quoted as 3.46% implying a forward price of USD 1,345. Assuming that the spot rate and
12-month LIBOR are unchanged then the implied lease rate is −0.46% (3%–3.46%). So
according to this approach no one is being paid to borrow gold, but rather a borrower
of USD is having to pay LIBOR + 0.46% (i.e. the swap rate of 3.46%) more than the
unsecured rate to borrow on a collateralised basis. Why would they do this? It would
suggest that they are unable to borrow at LIBOR and the rate of interest to borrow USD
offered by the gold swap is their best option.
4.5.2
Non-deliverable gold swaps
Non-deliverable forwards (NDFs) are well established in the foreign exchange market
and a similar product exists within the gold market. In an FX context, NDFs can be
used either as hedging or trading tools for situations where a market participant may
either be unwilling or unable to take delivery of a particular currency. As an example,
109
Gold
in a USDKRW contract at settlement then logically there should be an exchange of
US dollars against Korean Won. However, if this is done as a NDF then rather than
there being an exchange of cash flows, there is a single cash flow settlement in USD. In
the gold market these may also be referred to as ‘cash-settled swaps’ or ‘bullet swaps’.
Suppose a hedge fund client who does not wish to take physical delivery of the metal
agrees to a gold NDF for an amount of 10,000 ounces and is quoted a six-month forward
rate of USD 1,343. The value of the contract is therefore USD 13,430,000 vs. 10,000 oz.
of gold. The trader’s view is that the spot price of gold will not evolve as per the forward
rate, and at maturity will be less than the contract value.
Gold (XAU to use its currency code) is referred to as the reference currency, while
the USD leg is the settlement currency. The contract will be settled in USD at its maturity
in six months based on the spot value (‘settlement rate’) of gold at the time. The hedge
fund trader is designated to be the reference currency buyer. The relevant settlement
formula is:
USD notional amount × (1 − Forward rate∕settlement rate)
Suppose that in six months’ time the settlement rate is USD 1,320, then the settlement amount is:
USD 13,430,000 × (1 − 1, 343∕1, 320) = −USD 234,007.58
If the settlement amount is a positive number the reference currency buyer will
pay that amount in the settlement currency to the currency seller or if the settlement
amount is a negative number the reference currency seller will pay the absolute value
of that amount in the settlement currency to the reference currency buyer. In this case
the hedge fund trader profits as the rate moved as expected.
4.5.3
Deferred margin accounts
Deferred margin accounts are a type of derivative transaction in which a client can
express views on how the gold price is expected to evolve. The accounts work on a
leveraged basis but do not confer any rights of ownership on the physical metal.
Suppose a client wishes to take a bullish view on the evolution of the gold price and
agrees to a contract size of 10,000 oz. with the offering bank. The current gold price is
USD 1,300/oz. which implies an initial account balance of USD 13,000,000. However,
the client only partly finances this exposure and agrees to lodge an initial deposit of 8%
(USD 1,040,000) with the bank. The bank finances the balance of the required proceeds
(USD 11,960,000). The client is then allowed to make purchases or sales of the metal if
the following relationship is maintained:
Market value of metal − initial account balance + initial margin > 8% ∗
MAX (market value of metal, initial account balance)
(4.5)
The initial values are:
USD 13 m − USD 13 m + USD 1,040,000 = 8% ∗ MAX (USD 13 m, USD 13 m)
110
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
i.e. the equation is in balance.
These products are designed in such a way that the bank minimises its credit risk,
as typically they do not have any recourse back to the client. To illustrate how this relationship works assume that shortly after the contract is executed the price of gold falls
to USD 1,250. The left-hand side of equation 4.5 is therefore:
USD 12,500,000 − USD 13,000,000 + USD 1,040,000 = USD 540,000
The right-hand side of the equation is:
8% ∗ MAX (USD 12, 500,000, USD 13,000,000) = USD 1,040,000
Since the price of gold has fallen then the client will have incurred a loss of USD
500,000 (10,000 oz. x USD 50.00), and so their initial margin will have fallen in value.
The amount lent by the bank is still intact as the position could be unwound and
the bank’s principal would be repaid in full. In this case, the bank would require
the client to replenish the margin account such that the 8% threshold is maintained.
Some banks may impose a second threshold (e.g. 5%) below which the transaction
would be automatically unwound. This would protect the bank in the event of volatile
price moves.
If the price of gold were to rise to USD 1,350, then the relationship in equation (4.5)
becomes:
USD 13,500,000 − USD 13,000,000 + USD 1,040,000 = USD 1,540,000
Implying a profit of USD 500,000
However, the right-hand side of the equation is now
8% ∗ MAX (13,500,000, USD 13,000,000) = USD 1,080,000
Although the client would have to deposit some more margin, he would be allowed
to withdraw the ‘intrinsic value’ of the position, which is USD 460,000 (USD 1,540,000 −
USD 1,080,000). In this example the initial margin is retained in the account. He is not
able to withdraw any of the bank’s initial funding of the position and has no rights over
any physical gold.
From the offering bank’s perspective, they are exposed to a fall in the price of gold.
One of the ways in which they could hedge themselves is to sell gold futures. At the
time of writing, the contract specification for the gold future on the CME exchange is
for 100 troy ounces. The bank could therefore sell 100 futures as a possible hedge and
use the client’s margin to finance the exchange’s initial margin. This is not a perfect
hedge, though as the client’s margin account is referencing the spot price, the hedge
will reference a futures price. This leaves the bank exposed to movements in the ‘carry’
component of the future (i.e. LIBOR and lease rates).
111
Gold
4.6
YIELD ENHANCEMENT
Although central banks have had something of a love-hate relationship with gold, it is
still a fact that they hold considerable amounts of the metal. Since the metal does not
have any intrinsic source of income such as a dividend or a coupon, banks may seek
to earn a return on them by either leasing a portion of them to the market or executing a variety of strategies to enhance the yield. Lease rates are generally very low and
below interbank rates such as LIBOR. As such, in a rising interest rate environment,
it is tempting to suggest that banks should simply sell off their holdings for cash and
invest the proceeds!
One popular yield enhanced strategy involves selling a call option on a proportion
of the existing inventory of metal. This is commonly referred to as a ‘covered call strategy’. The sale of the call option will require the seller to deliver the underlying metal if
exercised. If the call is not exercised the central bank will simply collect the premium
and enjoy an enhanced return. A common misconception is that selling options while
holding the underlying asset is risk free. A simple example will demonstrate that this is
not the case.
Assume that the central bank decides to sell a 12-month call option based on a
portion of their physical holdings. We will assume that the current spot price is USD
1,300 per ounce; the three-month forward price is USD 1,303 and implied volatility at
20%. Using a simple option-pricing model the premium on an OTM European-style
option with a strike at USD 1,400 is USD 18.35 per ounce. Figure 4.10 illustrates the
concept with the analysis on a per ounce basis. The vertical axis is defined as the value of
the overall position in USD, while the horizontal axis represents the current spot price of
gold. The physical holding of gold is shown as a 45-degree line bisecting the horizontal
axis at the current spot rate. This position suggests that as the spot price of gold rises
so does the value of the underlying holding. The sale of the call option is shown with
an ‘at expiry’ payoff and the net position of the two transactions is also highlighted. As
the price of gold falls in the spot market, the value of the net position declines, but the
loss is offset by the income received on the option. Although the premium provides a
cushion for losses on the underlying holding, the overall position will show a loss.
If the price of gold were to fall, the net position starts to lose money beyond a spot
price of USD 1,281.65 (USD 1,300 − USD 18.35) so it is clearly not risk free. This is the
point where the cushion of the USD 18.35 received as a premium is eroded. However,
the central bank is still in a relatively better position than simply holding the metal. The
maximum value of the position is realised at a spot price equal to or greater than the
strike of the sold call option. Beyond this position the central bank loses on the call but
profits from the increased value of the physical gold. The net effect is that the profits
are maximised beyond this point. If the call is exercised against the central bank, they
do not need to deliver the metal and can choose to settle their obligations in cash.
Figure 4.10 shows that beyond USD 1,418.35 (USD 1,400 + USD 18.35), the returns
from holding just the physical gold would exceed that of the combined strategy. This is
a point where the call option breaks even and is equal to the strike plus the premium.
When considering the appropriateness of the strategy, the central bank must believe
that the price of gold will only rise by a relatively small amount.
112
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
15,000,000
Long gold position
10,000,000
Profit / Loss
Net exposure
Sale of OTM call
5,000,000
U/L Price
1,526.4000
1,507.5000
1,488.6000
1,469.7000
1,450.8000
1,431.9000
1,413.0000
1,394.1000
1,375.2000
1,356.3000
1,337.4000
1,318.5000
1,299.6000
1,280.7000
1,261.8000
1,242.9000
1,224.0000
1,205.1000
1,186.2000
-5,000,000
1,167.3000
0
-10,000,000
FIGURE 4.10 Covered call strategy.
From this analysis there are a few extra points of note:
▪ The bank could move the strike of the sold call option closer to the current underlying price. This would increase the premium of the option but also mean there is
a greater risk that the option will be exercised.
▪ The option was valued at an implied volatility of 20% with no reference to the state
of the skew. If the relationship between implied volatility and the strike resembled
a smile (e.g. Figure 2.7), then the position would be revalued using a higher implied
volatility, which would increase the premium.
▪ If the term structure of implied volatility for gold is upward sloping, then all other
things being equal, the bank would earn a higher premium if they were to trade a
longer-dated option.
4.7
SUMMARY
‘(one) feature that sets gold apart from other commodities and currencies is the
role that sentiment plays. Clearly it is important in all markets – the ability of
traders to ignore bad news and focus only on positive features that reinforce an
already-held view. However, in gold, this is taken to new levels. Why should a
piece of metal be a hedge against inflation? Why should it be seen as a store of
value? Is it desirable because it represents no one else’s debt? But then again,
the same is true for a lump of coal. The reason that gold has these attributes is
that we believe that it does. It has been embedded in human psychology and
reinforced over the millennia.’
—Spall (2010)
113
Gold
The key points from this chapter are:
▪ Gold traded on wholesale markets will conform to a set of agreed standards with
‘London Good Delivery’ bars being one example.
▪ The gold price moves on a constant basis although a twice daily auction formerly referred to as the ‘fixing’ establishes a benchmark value to aid transaction
settlement.
▪ There has been a shift in production away from the locations traditionally associated with gold (such as South Africa) to Asia (e.g. China).
▪ Central banks moved from being net sellers to net buyers as the price of gold
increased.
▪ Although the price of gold has been characterised as having strong price relationships with the US dollar and inflation, arguably a more compelling relationship
exists between the metal and real yields.
▪ The significant amount of metal held above ground, and the fact that most of the
gold ever mined is still in existence, has given rise to a leasing market.
▪ Given gold’s history as part of the monetary system it shares many similarities with
the currency markets.
▪ Some of the more unique swap products traded by the market include location and
quality swaps.
▪ Options on gold range from the vanilla to the exotic.
CHAPTER
5
Base Metals
5.1
OVERVIEW OF BASE METAL PRODUCTION
Typically, the mining process will begin with a geological survey of an area including
test drilling to determine the size of the deposit, its depth, and quality. From this a decision will be made as to whether or not the value of the deposit will be greater than the
cost to develop and operate the mine. This decision will consider several factors, some
of which may not be purely geological. These other considerations will include mining
infrastructure issues such as where to put the buildings or roads and will likely include
any related environmental issues. Lead times between the actual discovery and production can take many years and may often involve a substantial initial outlay. Many
banks wishing to lend to mining operations may often link the debt repayments to the
revenues generated and may involve some form of hedge to lock in future cash flows.
About 80% of all elements found in the earth are metal. However, they rarely occur
in a commonly recognisable form as they react to combine with other elements to create compounds. As a result, the production process will involve separating the metal
from the individual components with which it has bonded. Different metals will react
with other elements to varying extents. Silver and gold are very unreactive and do not
readily combine with any other element, and so are found in their pure state. At the
other extreme it is impossible to find francium in its pure form, as it will immediately
react with oxygen or water when exposed to the atmosphere.
Metals commonly occur in the form of compounds or ores where they are chemically bound to one or more elements. For example, the metal sodium is commonly
found as sodium chloride; sodium has bonded to chlorine. To produce and transform
the metals into a usable form the ore has to be separated from the host rock. The process
to remove the host rock is similar regardless of the type of metal concerned; the rock
and ore are ground down to a powder and are then separated mechanically.
The next stage in the production process involves the separation of the metal from
the compound. Breaking down the ore to form pure metal can be carried out by a number of different processes. The more reactive a metal is the harder it is to remove from
its ore and the more expensive the process. Several metal ores are broken down using
the addition of carbon and/or oxygen. Ores of very reactive metals have to be melted
down (smelted) and the metal removed by electrolysis.
Refining is the final stage of the production process where impurities are removed,
and the metal is transformed into a state that would allow fabrication of an end product.
114
Base Metals
5.2
THE COPPER LIFECYCLE1
5.2.1
Copper resources
115
Copper occurs naturally in the Earth’s crust in a variety of forms, and can be mined in
different ways such as surface or underground operations. Mine production in 2016 was
approximately 20 million tonnes, although overall production capacity was 10% higher.
One question that is commonly asked about many scarce commodities is ‘how
much is left?’ It is acknowledged that richer, low cost, long life ore bodies have already
been developed and are deteriorating quickly. Typically, the future availability of minerals is based on the concept of reserves and resources. According to the International
Copper Study Group (ICSG, 2017) ‘Reserves are deposits that have been discovered,
evaluated and assessed to be economically profitable to mine. Resources are far bigger
and include reserves, discovered deposits that are potentially profitable, and undiscovered deposits that are predicted based on preliminary geological surveys.’
The ICSG estimates that for 2016 ‘total copper resources’ amounted to 5,600 million tonnes. However, this is broken down into two main categories: undiscovered and
identified resources. Undiscovered resources are classified as those who ‘the existence
of which are only postulated’ and were estimated at 3,500 million tonnes. Identified
resources are defined as ‘resources whose location, grade, quality and quantity are
known or estimated from specific geological evidence. Identified resources include
economic, marginally economic and sub-economic components’. For 2016 these were
estimated at 2,100 million tonnes. Within the identified resources estimate sits copper
‘reserves’, which were estimated at about 720 million tonnes; reserves are: ‘That part
of the reserve base which could be economically extracted or produced at the time of
determination’. These reserves are determined so that they meet ‘specified minimum
physical and chemical criteria related to current mining and production practices’.
The question that follows is whether it is likely that the world will run out of copper.
Although copper is a finite resource, the level of reserves at any time usually tends to
be sufficient to meet about 40 years of demand. This can be explained by a number of
factors:
▪ As demand increases the price of the metal will rise meaning it will be more economical to extract marginal sources. Essentially, more ‘resources’ will become classified as ‘reserves’.
▪ Mining technology improves and so the estimates of resources and reserves evolve
over time.
▪ Copper is an example of a transformable rather than consumable metal and so can
be recycled repeatedly without any loss of performance.
▪ Innovation will change the way in which the metal can be used perhaps resulting
in economies.
1
Unless otherwise stated the data in this section was taken from The World Copper Factbook 2017
by the International Copper Study Group.
116
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
5.2.2
Uses of copper
The main applications of copper include:
▪ Electrical – power cables, appliances.
▪ Electronics and communications – data transmission, computer chips.
▪ Construction – external and internal applications (e.g. plumbing and window
frames).
▪ Transportation – motors, wiring, electric vehicles.
▪ Industrial machinery – gears, bearings, turbine blades.
▪ Other – coins, cookware, art.
5.2.3
The copper supply chain
A simplified overview of the supply chain for copper is presented in Figure 5.1. As with
many commodity supply chains some companies may be fully integrated or may just
offer a specific service.
The starting point to produce ‘primary’ copper is the mining of ore. This is the solid
material from which the metal can be extracted. The top five countries producing copper
are in descending order are:
▪ Chile
▪ Peru
▪ China
▪ United States
▪ Australia
The next step is to crush and grind the ore in order to remove all of the unwanted
rocky material. This is referred to as ‘concentrate’, which is comprised of several different elements typically 1/3 copper, 1/3 sulfur, and 1/3 iron silicate. It is possible for the
copper content at this stage to range between about 20–40%. The concentrate is sent
to a smelter, which transforms the material into ‘matte’ where the copper content will
increase to 50–70%. The top five countries smelting copper are in descending order:
Alloy Alloy
metals ingots
Concentrate
Matte
Ore
Mine
Blister
Anode
Wire Bars
Wire Rods
Cathodes
Smelter
Refinery
Semi
fabricator
Manufacturer
End of life
management
Scrap
Scrap
New scrap
New scrap
Old scrap
FIGURE 5.1 Copper supply chain.
Base Metals
117
▪ China
▪ Japan
▪ Chile
▪ Russian Federation
▪ India
The matte is then processed to produce ‘blister’ where the copper content is
98.5–99.5%. The blister is further refined and cast into anodes before the final refining
stage, which produces copper cathodes. A cathode is a 5–80 kg copper square and is
one of the main raw material inputs for ‘semi’ fabricators. Other refinery shapes may
include wire bars, ingots, billet slabs, and ‘cake’. The top five countries producing
refined copper are in descending order:
▪ China
▪ Chile
▪ Japan
▪ United States
▪ Russian Federation
Semi fabricators (‘semis’) convert the metal into several products and shapes such
as wire rod, tubes, and strips that can be used in the end manufacture of a specific
product. This part of the supply chain covers a wide variety of different entities such as
wire rod plants and brass mills. Brass is a mix of metals, referred to as an alloy, which
is comprised of copper and zinc. The top five countries producing refined copper are in
descending order:
▪ China
▪ United States
▪ Germany
▪ Japan
▪ Korean Republic
5.2.4
The role of scrap copper
The copper market distinguishes between ‘new’ and ‘old’ scrap metal. New scrap copper
is excess copper that is not incorporated into a final manufactured product. It is metal
that has been discarded as part of the fabrication or manufacturing process. ‘Old scrap’ is
metal that has been recovered from a manufactured product that has reached the end of
its life. This metal can then be re-introduced back into the supply chain. Indeed, refined
copper that has been made from scrap sources is termed ‘secondary copper production’.
5.2.5
Trading copper
In looking at the different elements of the supply chain, it becomes clear that where
copper is mined, smelted, refined, and fabricated could take place in many different
geographical locations and be performed by different market participants. The ICSG
argues that the major product categories traded internationally include:
118
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Concentrates,
▪ Blister and anode,
▪ Cathode and ingots,
▪ Scrap,
▪ Semis.
Typically concentrates and refined copper are the most imported and exported
elements.
In terms of the major trade flows of ores and concentrates (basically from mining
company to smelter) the major exporters are:
▪ Chile
▪ Peru
▪ Indonesia
The major importers are:
▪ China
▪ Japan
▪ Spain
In terms of blister and anodes (i.e. smelter to refiner) the major exporters are:
▪ Chile
▪ Bulgaria
▪ Namibia
Whereas the main importers are:
▪ China
▪ Belgium
▪ India
The main trade flows for refined copper (i.e. refinery to semis) include the following
main exporters:
▪ Chile
▪ Japan
▪ Russian Federation
While the main importers are:
▪ China
▪ Germany
▪ United States
Base Metals
119
In terms of semi-fabricated copper products, the main exporters are:
▪ Germany
▪ Taiwan
▪ China
While the main importers are:
▪ China
▪ United States
▪ Italy
5.3
ALUMINIUM
More aluminium is produced today than any other non-ferrous metal, with more
than 60 million tonnes produced each year. The prospects for the metal are generally
regarded as favourable as the demand for lighter, more durable, energy efficient, and
recyclable goods increase.
The starting point for the production of aluminium is bauxite. Once mined it is
crushed and dissolved in caustic soda at high temperatures and pressures. The solution contains dissolved aluminium ore, which can be separated from any un-dissolved
impurities, as they will have sunk to the bottom of the collection tank. Aluminium
ore is chemically referred as aluminium oxide (Al2 O3 ), but is perhaps most commonly
referred to as alumina2 and exists in the form of a white powder. The alumina is dissolved into a molten solution capable of conducting electricity. Two electrodes are then
attached, one positive and one negative, and the aluminium is extracted by electrolysis. Electrolysis uses electricity to separate the metal from its ore. As the electricity
passes through the molten solution, the aluminium is attracted to the negative electrode
and can be siphoned off to yield very pure aluminium. This entire extraction process is
energy intensive and as a result the cost of electricity will have a substantial impact on
production costs.
Four tonnes of bauxite yields about two tonnes of alumina, which yields one tonne
of aluminium.
Another important component of supply is the amount that is recycled. One of
the big advantages of producing aluminium via recycling is that the process consumes
considerably less electricity than primary production. The refining of aluminium is
renowned for consuming considerable amounts of electricity, which can account for
about 40% of production costs. It also explains why some smelters have chosen to locate
themselves in areas of relatively cheap energy costs such as Iceland. Recycling aluminium requires about 5% of the energy needed to make it from alumina.
2
As a rule of thumb alumina has traditionally been sold on a long-term contract basis with the
price being set at 12–14% of the price of aluminium.
120
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The largest producer (measured as smelting output) of aluminium is China
followed by the Gulf region and then North America and Western Europe. The largest
user of the metal is the transport sector, which accounts for about a third of demand.
A significant proportion of this is in the automotive industry for such things as engine
blocks and body panels. This would suggest that a downturn in demand in this
sector would have a significant impact on the demand for the metal. Construction is
the second largest user of the metal (25%) with packaging constituting a significant
proportion (16%).
5.4
THE STEEL MARKET3
By way of contrast to the copper market, steel is not a commodity that exists naturally
in the ground. Steel is an alloy of iron and carbon and contains small amounts of other
elements such as manganese and silicon.
Iron ore is first reduced to form molten iron, sometimes referred to as ‘pig’ iron. This
molten iron is then transferred to a basic oxygen furnace (BOF; about 72% of all production) where it is combined with a proportion of scrap steel and once various impurities
are removed, the resulting product is molten steel.
Alternatively, it is also possible to produce steel in an electric arc furnace (EAF;
about 28% of all production). Arguably, the key differences between the processes relates
to the production inputs. The EAF process tends to use a greater proportion of steel
scrap. Steel can be infinitely recycled without loss of quality.
To produce 1,000 kg of crude steel using the BOF technique, the main required
inputs are:
▪ 1,370 kg of iron ore,
▪ 780 kg of coal,
▪ 270 kg of limestone,
▪ 125 kg of steel scrap.
To produce 1,000 kg of crude steel the EAF process requires about:
▪ 710 kg of steel scrap,
▪ 586 kg of iron ore,
▪ 88 kg of limestone,
▪ 2.3 GJ of electricity.
The main by-product of steel production at this point in the supply chain is ‘slag’
which has application in areas such as cement production.
Steel is often thought of as a single commodity, but there are more than 3,500 different grades of steel, significant portions of which have been developed within the last
3
The facts in this section have been primarily sourced from the World Steel Association
(https://www.worldsteel.org/steel-by-topic/statistics.html)
Base Metals
121
20 years. ‘Unalloyed’ carbon steel (e.g. just iron and carbon) comprises the majority
of steel production. The addition of other metals will improve the quality of the steel.
For example, adding manganese will make it harder while nickel and molybdenum
improves its resistance. Coating steel with zinc or nickel will help prevent rusting.
The next stage of the process is to cast the molten steel into semi-finished products.
These comprise of ‘long products’ (e.g. bloom and billet, mostly produced from a BOF)
or ‘flat’ products (e.g. slabs, mostly produced from an EAF). These products can then
be rolled into different shapes.
Semi-finished long products are rolled into:
▪ Wire rod – used in construction transport and energy.
▪ Rebar – reinforcing steel bar used in construction.
▪ Sections – used in construction.
▪ Structural steel and rails – used in construction.
The flat products are rolled into:
▪ Plate – used in construction, appliances, engineering, and energy.
▪ Hot/cold rolled coil – used in construction, transport, and appliances.
The key applications of steel are:
▪ Buildings and infrastructure – 51%
▪ Mechanical equipment – 15%
▪ Automotive – 12%
▪ Metal products – 11%
▪ Other transport – 5%
▪ Domestic appliances – 3%
▪ Electrical equipment – 3%
5.4.1
Factors impacting the price of steel
Since the raw material inputs in the production process will often be traded markets,
it is no surprise that they will have a significant impact on the price of steel. The key
inputs are:
▪ Metallurgical coke (a refined carbon product made from coal).
▪ Iron ore,
▪ Scrap steel,
▪ Energy (e.g. coal and natural gas),
▪ Freight costs associated with movement of raw materials.
On the demand side, the main consumer of steel is China, followed by India and
the Middle East. Notably, China is also the largest producer of steel.
122
5.4.2
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Steel risk management
Although the market for steel is sometimes considered to be the second largest after
crude oil, futures contracts are a relatively recent addition to the suite of risk management products.
At the time of writing, the London Metal Exchange (LME) offers two cash-settled
steel futures; one references steel rebar and the other is scrap steel, the latter being
the more liquid of the two contracts. The contract specification for the scrap steel contract is shown in Table 5.1.
TABLE 5.1 Scrap steel futures contract.
Contract type
Futures
Deliver type
Lot size
Contract period
Price quotation
Final settlement price
Cash-settled
10 tonnes
Monthly out to 15 months
US dollars per tonne
Based on the monthly average price index price of the “Platts
TSI HMS 1/2 *0: 20 CFR Turkey” assessment
Source: Based on LME, London Metal Exchange- Steel Scarp, 2020.
The contract reflects the fact that Turkey plays a central role in the market for
steel scrap. One notable difference of this contract compared to the normal conventions
applied to LME contracts is that it is cash, rather than physically settled, referencing a
Platts index.
Risk management products have also emerged for iron ore and coal, which will be
considered in subsequent chapters.
5.5
THE LONDON METAL EXCHANGE
One common aspect of many commodity markets is the struggle to establish a
single common and transparent pricing basis. For those participants involved in
negotiating commercial contracts it may be difficult to determine if they are paying a
“fair price”. In addition, the settlement of derivative transactions requires agreement
on a single common price for the underlying asset.
One of the main attractions of the London Metal Exchange (LME) is that their
traded prices can provide the basis for a substantial proportion of such commercial
transactions. The LME (2017) point out that the exchange is designed with physical
market participants in mind (e.g. producers, processors, and consumers) who can avail
themselves of the different risk management products that are available and can also
use the exchange as a possible source of supply. However, this type of entity makes up
about 1/3 of the total volume, with a variety of different financial market participants
making up the balance.
The LME offers futures and options that will allow participants in these markets to
manage their price risk. Although the LME has many features in common with other
Base Metals
123
financial exchanges (e.g. standardised contracts) the design of its forward price-based
contracts often has more in common with over the counter (OTC) style instruments.
It should be noted that a significant proportion of base metal transactions are executed on an over the counter basis. If a corporate customer wished to execute a transaction which is exactly tailored to their needs, the OTC market offers greater flexibility
and a wider range of products. However, the LME offers an efficient mechanism where
the risk taken in an OTC contract can be hedged using their suite of risk management
products.
5.5.1
Exchange traded metal futures
Futures contracts represent the commitment to either buy or sell a particular metal at
a pre-agreed price at a specified date in the future. LME transaction will require both
parties to the trade to physically deliver the metal if the contract is held to maturity.
Although it is possible to use the futures contract as a source of supply, it is more likely
that the contract will be used as a financial hedge. Using the future in this fashion permits the user to separate the supply of the metal from the associated price risk. In reality,
only a very small proportion of futures trades will go through to final settlement, typically about 1%4 . Physical settlement can be avoided by taking out an equal and opposite
trade for the same delivery date (termed the ‘prompt’ date). The two transactions will
offset each other, and the counterparty will cash settle the trade paying or receiving the
difference between the two prices.
The exchange currently offers futures contracts on four different categories of
metals:
Non-ferrous
▪ Copper
▪ Primary aluminium
▪ Aluminium alloy
▪ North American Special Aluminium Alloy (NASAAC)
▪ Lead
▪ Nickel
▪ Tin
▪ Zinc
Ferrous
▪ Steel scrap
▪ Steel rebar
Minor metals
▪ Cobalt
▪ Molybdenum
4
Source: LME (2017)
124
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Precious metals
▪ Gold
▪ Silver
5.5.2
Exchange traded metal options
The exchange offers options on the underlying futures contracts, so the buyer has the
right but not the obligation to buy or sell an LME futures contract at a predetermined
price at a pre-agreed time in the future. The exchange also offers Traded Average
Price Options (or TAPOs). Average price options (sometimes referred to as ‘Asian
style options’) are options where the settlement of the contract is not made against
a single market price but against an average for a predetermined period. They are
attractive from a buyer’s point of view as the premium cost is lower than an equivalent
non-averaging option. TAPOs are settled against the LME Monthly Average Settlement
Price (MASP), which is a figure published by the exchange that represents the average
of all the daily settlement prices for a particular month.
5.5.3
LME prices and contract specification
In Section 5.2 it was argued that the supply chain for copper could be broken up into
different components:
▪ Mining (ores and concentrates),
▪ Refining,
▪ Semi-fabrication,
▪ End product.
One of the issues central to the exchange traded metals markets is the exact form of
the metal to be physically delivered at maturity. To ensure the contracts are standardised and are useful to market participants, the exchange sets specifications for form,
quality, and shape, which are most widely accepted by the different market users. The
LME contracts are usually for metals that can be easily transformed into an end product by a commercial user. As such the LME contract is designed to reflect the form and
associated price of the metal at the refining stage of production. As a result, the LME
price can be used as the basis for pricing a wide variety of commercial metal contracts
where the actual metal content may vary. For commercial contracts based on metal in
a state different to that specified in the LME contract, a mutually agreeable adjustment
to the LME price could be applied. This may be a discount (e.g. ores and concentrates)
or a premium (e.g. semi-fabrication or end product users).
One example of how this is applied can be illustrated by considering an intermediate pricing system that is used between miners and smelters referred to as ‘TC/RCs’. TC
stands for treatment charges (i.e. how much it costs to smelt the copper concentrate)
while RC stands for refining costs. Theoretically it represents the cost of converting a
tonne of concentrate into metal. They are charged by a smelter to a mine and as such
represent revenue for a smelter and cost to a mine.
Suppose a mine delivers concentrate to a smelter with 30% copper content. The
price paid for the concentrate would reference the LME price but at a slight discount
125
Base Metals
(e.g. 96.5% of the current market price). If the LME price is quoted as USD 7,000/tonne,
the amount payable by the smelter would be USD 2,030 (30% x 96.65% x USD 7,000).
However, this initial payable amount is reduced by the TC/RCs. The treatment charge
can be quoted as a USD amount per ‘dry metric tonne’ (DMT) of concentrate. The
refining charge is quoted differently as it is based on the quantity of copper within the
concentrate; it may also by convention be quoted in US cents per pound. So if the copper content is 30% and the refining charge is, say, USD 0.045/pound of copper content,
then the refining charge would be 30% x USD 0.045 x 22,046 = USD 22.8/DMT (there
are 22,046 pounds to a metric tonne).
The exchange also stipulates common standards for weights and strapping. Participants trade in lots rather than tonnes, with aluminium, copper, lead and zinc being
sized at 25 tonnes. The NASAAC and aluminium alloy contract is for 20 tonnes, with
nickel at six tonnes and tin at five tonnes.
For example, the specification for the copper futures contracts is outlined in
Table 5.2.
The contract specification for copper options is outlined in Table 5.3.
The expiry structure of LME contracts is relatively involved and merits a few comments. Metal can be traded for settlement on any business day three months into the
future. For contracts expiring between three and six months, settlement is then weekly
with Wednesday being the designated expiry date. Depending on the specific contract,
settlement beyond six months will take place monthly out to a maximum of 123 months,
each settling on the third Wednesday of the month. As a rule of thumb contracts for
three-month delivery are the most liquid; this is based on the historic time it took to
transport metal from its origin to London.
5.5.4
Trading
Trading at the LME is by a mix of open outcry and electronic transactions. The focus
of open outcry trading is the ‘ring’ sessions. There is a morning and an afternoon session, which follow the same general structure. Each metal is traded twice in designated
five-minute sessions and at the end of the second ring in the morning session, LME
TABLE 5.2 Copper futures contract specification.
Contract
Lot size
Form
Delivery (‘prompt’) dates
Quotation
Settlement type
Clearable currencies
Grade A copper conforming to the chemical composition
stipulated by the exchange. All copper deliverable against
an LME contract must be of an approved brand.
25 tonnes
Cathodes
Daily from cash to three months. Then weekly every
Wednesday from three to six months. Then monthly every
third Wednesday from seven months out to 123 months.
US dollars per tonne
Physical
US dollar, Japanese yen, sterling, euro
* = Prompt date refers to the day that the physical metal is exchanged for cash.
Source: Modified from LME, London Metal Exchange- Copper, Future Contract Specifiation, 2020.
126
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 5.3 Copper options contract specification.
Underlying contract
Lot size
Price quotation
Option style
Delivery dates
Settlement type
Expiry date
Premium quotation
Strike price
LME copper futures – third Wednesday prompt of the contract
month.
25 tonnes
US dollars per tonne
American
Monthly from the first month out to 63 months
Physical
Any LME business day up to and including the first Wednesday of
the expiring option month.
US dollars per tonne
USD 25 gradations for strikes from USD 25.00 to USD 9,975
USD 50.00 gradations for strikes from USD 10,000 to USD 19,950
USD 100.00 gradations for all strikes over USD 20,000
Source: Data from LME, Copper options contract specification, The London Metal Exchange. www.lme.com
staff monitoring activity within the ring will determine the ‘official’ closing prices for a
variety of different maturities. Although it varies between metals, for copper the official
prices quoted are for:
▪ Cash (i.e spot value)
▪ Three months
▪ Three forward December prompts
This official fixing allows the market to establish a transparent benchmark against
which trades can be settled.
For example, for copper the ring trading times are:
First session
First ring
Second ring (official)
Kerb trading
12:00–12:05
12:30–12:35
13:25–13:35
Second session
Third ring
Fourth ring
Kerb trading
15:10–15:15
15:50–15:55
16:20–17:00
Kerb sessions run after the main ring sessions and during this time, all the metals
are traded with more than one trader per company allowed within the ring. In addition to the ring sessions it is possible to buy and sell metals around the clock on the
exchange’s electronic trading application.
Base Metals
5.5.5
127
Clearing and settlement
Settlement describes the point in time where an asset is exchanged for cash. Clearing
relates to all those activities that happen after trade execution but before settlement. It
is quite common for people to use the terms ‘clearing’ and ‘settlement’ interchangeably.
One of the features common to all exchanges is the existence of a central clearing
house and for the LME these functions are performed by its own entity LME Clear.
Its primary role is to ensure that all purchases and sales executed on the exchange
are settled in a timely manner. Once a trade is executed, both sides to the deal input
the transactions to a common system and if the details agree, the trade is considered
matched.
As a result, the clearing house then becomes the legal counterparty to both sides of
the transaction. That is, it becomes the seller to the buyer and the buyer to the seller.
Neither of the original counterparties will have any further contractual commitments
to each other, as the clearing house is now their counterparty. With the clearing house
acting as the counterparty to both sides, the risk of the original counterparty defaulting
is transferred to the clearing house. If one of the original contracting entities were to
default, then the exchange would step in and guarantee the financial settlement with the
non-defaulting entity. However, the probability of the clearing house defaulting is very
slim as these entities are usually very well capitalised. It should be noted that this mitigation of credit risk only extends to those transactions executed directly on the exchange.
A client who asks a broker to act on their behalf is still exposed to the full credit risk of
their broker and vice versa.
Even though the likelihood of the clearing house defaulting is slim, one of the techniques used to further mitigate the risk is the margining system. Margin is collateral that
is collected at the start of the transaction (‘initial margin’) or as the value of the contract
changes (‘variation margin’). Although cash is the most popular form of collateral, it
may also be possible to deposit risk free securities (e.g. government bonds). In addition
to the formal margining system covering transactions executed in the ring (or electronically), institutions trading on behalf of clients will also margin customers in order to
mirror the payments they will make to the clearing house.
The margining system on the LME works in a different manner than that seen on
the various financial exchanges. On financial exchanges, variation margin is calculated
based on the change from the previous day’s closing price. The margins are paid or
received daily5 over the life of the deal, with the final settlement price – the exchange
delivery settlement price (EDSP) – occurring on the last trading day. The impact of this
margining system is to effectively make the future a series of daily contracts, which are
instantaneously closed out and reopened at the closing daily futures price.
The following examples will illustrate the concept. We will assume that a ring member has executed a trade where they buy one lot of aluminium at USD 1,900 per tonne
for delivery in three months’ time. They will have to pay an initial margin, which is
assumed to be USD 100.00. Suppose that one day later the price for delivery on the
original prompt date is now USD 1,950.
5
Typically margins are calculated based on the closing price of the metal and are payable first
thing the following day.
128
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
In the first scenario we will assume that the member closes out the trade at this
point. He has bought at USD 1,900 and sold one day later at USD 1,950. Assuming no
further trades, the initial margin of USD 100.00 is returned as is the USD 50.00 profit
which is their variation margin.
In the second scenario we will again assume that the price has risen but that the
trader decides to retain the position. Even though his position has increased in value,
he receives no cash benefit. Unlike many financial exchanges the LME will only pay
profitable variation margin when the position is terminated. However, if the price of
the metal were to fall after one day to USD 1,850, the trader would be required to make
immediate cash payment to the clearing house to support this loss. Losses will need to
be made good immediately (like financial exchanges) whereas profits will not be paid
until maturity (like an OTC contract).
Since the financial crisis, the differences between OTC and exchange-based transactions have begun to blur and continue to evolve. Most financial markets are evolving
towards three different structures:
▪ OTC traded, OTC cleared and settled,
▪ Exchange traded, exchange cleared and settled,
▪ OTC traded, exchange cleared and settled.
For OTC transactions that do not involve the exchange, the credit risk between broker and client could be managed in a variety of ways. Some institutions may allow profits
and losses on contracts to be handled as part of some agreed credit facility. Other institutions may operate a system where all losses and profits are remitted as and when they
occur, or they may simply mirror the LME process. However, a very common method
of managing the credit risk is through the taking of collateral. This is often governed by
the terms of ISDA documentation (International Swaps and Derivatives Association).
The part of the documentation that refers to collateral is referred to the Credit Support
Annex (CSA). These cover:
▪ How often collateral can be taken.
▪ The point at which collateral will be taken (similar in some respects to an overdraft
limit).
▪ The minimum amount that can be transferred between the two counterparties (for
example £100,000).
▪ The type of acceptable collateral (e.g. cash and certain government securities).
5.5.6
Delivery
Settlement of futures contracts on the LME takes place in a unique fashion in comparison to a financial exchange. On financial exchanges, instruments can be bought and
sold at any time although they will typically have fixed maturity dates set at three-month
intervals. LME contracts can be traded for settlement on any chosen day (the ‘prompt
date’) for up to three months from the trade date, then weekly for maturities between
three and six months followed by monthly settlement and then monthly settlement out
Base Metals
129
to 15, 27, 63, or 123 months forward. This makes the LME future more like a centrally
cleared over-the-counter forward.
Most LME contracts require physical settlement if they are held to maturity. As
mentioned previously, the percentage of contracts going to final delivery is relatively
small, but nonetheless the exchange can be used as the ‘market of last resort’ to procure
supplies. To facilitate the potential delivery of the metal the LME has approved over
600 warehouses in 35 different countries typically sited in locations of net consumption rather than production. The exchange does not own or operate the warehouses
and neither does it own the metal that is held therein. However, market analysts watch
the volume of metal held within the warehouses as a gauge of currently available
inventories.
When an institution agrees to sell metal on the LME, it may choose the warehouse
destination where delivery will take place. The contract specification outlines what can
be delivered and the exchange has strict requirements on quality, shape, and weight.
At the time of writing, the LME lists over 90 different brands of copper that can be
delivered to satisfy a short position. If the seller delivers metal to the warehouse it will
receive a warrant, which acts as evidence of ownership. The deposit of metal into the
warehouse need not be in support of an LME contract and the issue of a LME warrant
will only take place if the metal conforms to the contract standards outlined by the
exchange. The terminology is that the metal is ‘put on warrant’. If the owner of the metal
on warrant wishes to remove the metal, then the warrant will need to be canceled so it
becomes ‘off warrant’. The right of a buyer to take delivery of the metal at the expiry of
a futures contract is conferred by the transfer of a warrant from the seller. The issue of
warrants is done by the warehouse in the location chosen by the seller for delivery, but
the LME has overcome the administrative difficulties of managing a physical warrant
system by introducing an electronic transfer system (LMEsword). This acts as a central
depository for all LME warrants, which are issued in a standardised format and can now
be transferred rapidly between buyer and seller.
In certain financial futures markets (e.g. bond futures) the contract seller has a
variety of embedded delivery options. In the metals market the seller can choose the
location of delivery and the brand. If a seller is in possession of a number of warrants
and is faced with having to deliver the metal, then logically they will choose the location that is most favourable to them (so probably less favourable to the buyer) and the
brand that is least attractive. In financial futures markets this feature is referred to as ‘the
cheapest to deliver’ option. This means that although there is a single ‘global’ futures
price for all metal deliveries, the quoted price will reflect this optionality. As a result,
there is a market to ‘swap’ warrants, which takes place off the exchange. So, if the terms
of the warrant are not suitable for the buyer, they may seek to swap the warrant for a
more favourable type. This has led to the development for premiums for warrants that
confer delivery in a popular location or for a popular brand of metal. LMEsword allocates warrants randomly to those buyers with outstanding positions. This means that a
buyer does not know which brand and into which location they will be delivered.
On settlement date, the exchange will analyse the position of each individual entity
netting off as many of their remaining purchases and sales, with any price differences
settled in cash. The remaining net position will then go to settlement. Since the position
on the exchange is always in balance (that is the total number of purchases or ‘longs’
130
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
equals the total number of sales or ‘shorts’) at expiry, the exchange will take up the
warrants from the market longs and will randomly assign them to the market shorts.
The actual cash settlement amount may differ from the forward price agreed under the
terms of the original deal because each contract allows the actual weight of the metal to
differ by +/− 2% of the contract standard. As a result, the final price paid by the buyer
will need to be adjusted to reflect the actual amount delivered.
5.6
BASE METAL PRICE DRIVERS
The main factors that drive the price of base metals markets include:
Raw material availability – richer, low cost, long life ore sources have already been
developed and are deteriorating quickly.
Level of inventories – in the gold chapter it was highlighted that the bulk of all known
gold was held above ground and as a result the forward curve was generally in contango.
However, it is different for base metals where the balance of above and below ground
supply is the opposite. For example, copper has traditionally suffered from relatively low
above-ground inventories and as such its forward price curve can be prone to extended
periods of backwardation.
Government fiscal and monetary policy – at a simple level if the government were to
implement an expansionary economic policy, this should stimulate economic activity
and increase the demand for resources such as base metals. For example, in the copper
market such factors as the level of car output, semiconductor demand, and new housing
activity all influence the price of the metal.
Exchange rates – base metals are traded in US dollars and so movements in the
exchange rate will have an impact on the balance of demand and supply. If the dollar
weakens relative to other currencies, the domestic currency cost of buying the metal
will decrease, leading to an increase in demand and therefore an increase in the USD
price.
Chinese and Indian demand – the industrialisation of the Chinese economy and the
rapid development of India has made the two countries substantial importers of many
metals due to the lack of their own natural resources. This has contributed substantial
upward pressure to the price of many commodities including base metals. For example,
in 2016, copper production was estimated6 at 23,339 thousand tonnes, (abbreviated here
to ‘kt.’) of which China was the largest single country producer at 8,436 kt. However,
China consumed 11,446 kt. which represented about 50% of global demand. To put this
into some context, European demand was 3,700 kt. while North America accounted for
2,336 kt.
Availability of finance – after the financial crisis of 2007/2008 many banks were
unwilling to extend credit and so as a result capital expenditure declined. In theory, this
could eventually feed through to lower production and higher prices (all other things
being equal).
6
Barclays Bank (2018) ‘Copper at the crossroads’.
Base Metals
131
Capital spending and exploration – during periods of sustained low metals prices,
mining companies are reluctant to invest in their own infrastructure or instigate new
exploration projects. However, as the prices of metals rise, it will then prove difficult to
bring new production on stream to benefit from rising prices. With many commodity
markets there may be a substantial delay until new supplies of metal can be brought
to market during which time the excess demand will cause the price of the metal to
rise further. For example, it could take up to 10 years to produce metal from the point
of its discovery. Over the years this production lag has increased due to factors such as
increased construction costs due to rises in the price of metals (such as steel) greater
labour and environmental regulations and a weaker US dollar. Although the argument
has been made from a mining perspective, similar arguments could be made with
respect to smelting and refining capacity.
Environmental approvals – these can often be very lengthy processes and may
encourage producers to focus on those jurisdictions where the standards are relatively
loose.
Substitution – as commodity prices rise the possibility of product substitution
increases. With respect to base metals this may not be possible in the short term,
but over the course of, say, two to three years this may well become feasible as
manufacturers develop products with a smaller metal content.
Production disruption – this usually manifests itself in two ways. The first would be
the cessation of production due to labour disputes. Although rising prices may boost a
producer’s profitability, it may lead to an increase in wage demands from the work force
and possible industrial action. Barclays (2018) estimates that labour costs account for
26% of the cost of producing copper. The second reason relates to industrial accidents
such as landslides, which may halt production for a period.
Quality of infrastructure – although it may be tempting for producers to seek out
relatively cheap sources of supply, one of the consequences may be that the state of
local infrastructure may present challenges in terms of extraction and onward movement along the supply chain. If the extraction of the metal takes place in an emerging
economy, there may also be issues surrounding the provision of a constant supply of
energy. For example, in 2007, China agreed to build or refurbish Congolese infrastructure to a total of USD 12 billion in exchange for the right to mine copper of an equivalent
value7 .
Rising nationalism and protectionism of domestic resources – this relates to the fact
that the extraction of the ore may be undertaken by a private foreign entity. A change
in political climate may result in a demand to ‘take back’ the resources into local
ownership.
Production costs – one of the features of the general rise of commodity prices is the
impact it then has on the costs of subsequent production. Operating costs could rise
due to factors such as higher steel prices and energy costs. For example, the impact of
higher energy prices is perhaps most apparent in the aluminium market where 50% of
the production costs arise from these charges.
7
Economist (2008), ‘A ravenous dragon’.
132
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Investment demand – the equity bear markets of the late twentieth/early twentyfirst century led many institutional investors to seek out alternative asset classes
to enhance their return. Evidence suggests that returns on commodities tend to be
negatively correlated with those on financial assets, offering a powerful incentive for
investors to diversify. Other investment motivations include the wish to find a hedge
against inflation and the desire to benefit from the rise in prices in a bull market.
One of the most common indicators of this activity is the ‘Commitments of
Traders’ report produced by the Commodity Futures Trading Commission in the
USA. This report shows the positions (in the US markets) for ‘commercial’ and
‘non-commercial’ users. Commercial users are defined as those entities that would
use futures for hedging some form of underlying exposure. Investment demand
would therefore be captured within the non-commercial users’ category.
The impact of the business cycle – here the focus is essentially on the level of metal
inventories. Figure 5.2 illustrates in a generalised fashion the impact of the business
cycle on the base metals market. In this very simple model, the business cycle is represented as having four stages. Every economy will progress through each of these four
stages in order but the length of each stage (how long it lasts) and the depth (how severe
it will be) will be both different and unpredictable.
In phase I (‘prosperity’) the economy is expanding and prices in general are rising. As economic activity increases there is an incentive to produce more and metal
inventories tend to be relatively full. Available production capacity can catch up with
the increase in demand. In phase II (‘transition’) asset prices overshoot their ‘fair value’
and demand starts to fall. As a result of the general decline in demand, inventories are
wound down. This is because consumers are reluctant to buy new metal at high prices
and so meet their requirements by running down existing inventories. There will also
be an incentive to recycle the metal from scrap products as it increases in value. At
this point, financial investors will start to withdraw their funds. Phase III (‘Recession’)
represents a bear market, as demand is very low. Here there is very little incentive to
produce the metals given the combination of lower prices and lower demand. In phase
IV (‘Recovery’), growth accelerates as businesses restock metal inventories due to rising
prices and demand. At this point inventories may tend to fall, and metal prices rise, as
the increasing demand cannot easily be met out from available productive capacity.
Prosperity
I
Transition
II
Recession
III
FIGURE 5.2 The impact of the business cycle on base metals.
Recovery
IV
Base Metals
5.7
133
ELECTRIC VEHICLES
Broadly speaking there are three main types of electric vehicle (EV):
A hybrid vehicle uses two systems to power the car. There is a traditional internal
combustion engine that uses gasoline and an electric motor. The electric motor could
be used to provide additional acceleration but can also be used to completely power
the car at low speeds and short distances. The electric motor is charged by the internal
combustion engine as well as the braking system.
Plug-in hybrid electric vehicles will have some similarities to hybrid vehicles except
that the battery capacity is greater allowing them to run on this source of energy for
longer times and for greater distances. The battery is recharged from mains electricity.
A full battery electric vehicle is only powered by an electric motor that derives its
energy from batteries. These batteries can be charged by plugging the car into mains
electricity.
One of the key arguments made in favour of EVs is that they do not produce emissions, but this tends to ignore the fact that to generate electricity then either coal or
natural gas will be burnt. This simply moves the emissions issue from the vehicle to
another part of the supply chain.
From a commodities perspective, electric vehicles will impact several aspects of this
text for a variety of reasons:
▪ Will the advancement of battery technology see new traded markets develop?
▪ How will the metals currently used in car production evolve? For example, palladium and rhodium are not typically used in the EV production process.
▪ How will the move to electric motors impact the main energy markets of oil, refined
products, natural gas, and electricity?
▪ How will the energy infrastructure develop to accommodate the increase in
demand for electricity?
As of the time of writing the LME currently offers EV risk management products
in cobalt, nickel, copper, and aluminium. To meet the increasing demand for EVs then
new products likely to be developed include lithium, nickel, sulfate, and manganese.
5.7.1
Lithium
Most electronic devices use lithium-ion batteries. Lithium has been described as the
‘new gasoline’ given its importance in EV batteries. Although the batteries may contain
40–80 kg of lithium, scientists are always looking at ways to reduce cost and boost power
given that many consumers cite ‘range anxiety’ as one of the reasons they are reluctant
to invest in an electric car. This could feasibly mean an increase in alternative cheaper
materials as substitution takes place.
The supply chain for lithium shares common characteristic with those of other
metals:
▪ Mining
▪ Concentration/semi-processing
134
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Chemical refining
▪ Battery precursors
▪ Battery cathode materials
▪ Li-ion batteries
5.7.2
Cobalt
This metal is used as a compound with lithium in the cathode of lithium-ion batteries. It is produced mainly as a by-product of copper and one of the largest suppliers
is the Democratic Republic of Congo, who sells a substantial proportion of the output
to China for processing into battery materials. The metal has traditionally traded at a
relatively high price and so there will be an incentive to find cheaper alternatives. In
2017, Volkswagen issued a tender to obtain a minimum of five years of cobalt supply
at a fixed price. However, in anticipation of potential increasing demand and therefore
rising prices, miners were reluctant to sign up.
5.8
STRUCTURE OF MARKET PRICES
5.8.1
Long-term prices
‘Long-term prices’ is a commonly used phrase in commodities and is worthy of some
analysis. In theory, a long-term price is one that ensures a balanced market and stable
prices over the medium to long term. It is often one that is used for longer-term valuation and investment analysis. So, if a bank is considering lending to a producer to
finance a new mine operation, they may insist that as part of the transaction, a proportion of the production should be sold on a forward basis. If the new mine was not going
to see any production for five to six years, the analyst is faced with making predictions
about the forward price at some future date rather than today’s forward price. The big
question is – which price should be used?
Arguably, there are perhaps five different approaches: 8
▪ Mean reversion price – an average over an extended history of prices (and whether
it is adjusted for the impact of inflation).
▪ Future average price – the average of annual forecasts over an extended future
period.
▪ Equilibrium price – a qualitative assumption or derived from some perceived relationship between price and levels of inventories.
▪ Forward market long-term price – using the back end of the current forward curve.
▪ Incentive price – the price needed to encourage supply to meet future demand. The
theory is based on the notion that producers will only invest if they believe that
prices will be high enough to cover all their investment costs and provide an adequate return on capital over the life of the project.
8
Barclays (2004), ‘Copper market outlook – The twin peak’.
Base Metals
5.8.2
135
How do forward curves move?
When prices increase it is the shorter end of the forward curve that exhibits greater
volatility. The shorter end of the forward curve will be more influenced by incidents
in the physical market, whereas long-term prices are determined by mostly financial
considerations (e.g. inflation and interest rates). An example of this was illustrated with
respect to crude oil in Chapter 2 (Figure 2.2).
This suggests that there are three main ways in which a forward curve can move:
▪ Directional changes – the curve moves up or down in a broadly parallel manner.
▪ Slope changes – the slope of a curve measures the difference between two prices at
different maturities. So, between any two maturities it is possible for the curve to
either steepen or flatten.
▪ Shape changes – this relates to the degree of curvature and is measured as the difference between any three points. Although there are several ways in which this could
be calculated, it is usually measured as a central maturity minus a shorter-dated
maturity minus a longer-dated maturity.
5.8.3
Are forward prices forecasts?
This was discussed at length in Chapter 2 and perhaps can be summarised by a simple
statement: ‘it depends’. There is something of a market myth that forward prices are
somehow the market’s ‘best guess’ as to future spot levels (i.e. today’s three-month forward price is a forecast of the spot price in three months) but this supposes that price
formation is solely determined by expectations, which is a very severe constraint.
The best way to think of a forward is to view it as a benchmark rate against which
hedging decisions could be assessed. Suppose that the ‘cash’ (i.e. spot) price of copper
is USD 6,700 and the three-month rate is USD 6,800. A market participant looking to
sell the product in three months’ time could either lock into a price of USD 6,800 or do
nothing and hope that over the period prices will rise above the initial forward rate. If
the seller believed that the cash price was going to be higher than the initial forward,
then the ‘do nothing’ scenario is the optimum solution. If they believed that cash was
going to end up below the initial forward rate, then they should sell forward. If they felt
that the cash price would evolve in line with the forward market and end up at USD
6,800 then they would be indifferent between the two choices9 .
As the price of the metal rises there is a possibility that the market could move into
backwardation. Indeed, as the price rises, the slope of the forward curve becomes more
steeply backwardated. In these conditions (high prices and steep backwardation), producers will be less tempted to sell their production forward. Anecdotally, at very high
price levels, producers are often less keen to lock into attractive prices. This is often
a result of shareholder pressure that wants the company to benefit from rising prices.
It may also be simple over-optimism. Somewhat paradoxically, anecdotal evidence suggests that producers tend to be more active hedgers in low price environments. In addition, consumers may decide to take some advantage of the cheaper forward purchase
9
This simple analysis ignores other hedging strategies such as options.
136
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
price for the metal. Investors may wish to take advantage of the rising price environment and take exposure to the market using exchange traded futures, further adding to
the longer-dated buying pressure. These actions will, to an extent, act as a brake to the
degree of backwardation experienced by the market.
5.8.4
The role of marginal costs
The classic definition of marginal cost is the additional cost of producing one additional
unit of a particular product. Within the base metals world discussions on how long-term
prices are expected to evolve often centre around this measure and is usually shown
diagrammatically.
A hypothetical marginal cost curve is shown in Figure 5.3.
The graph plots the production capacity and costs for an entire sector (e.g. copper
or coal). The horizontal axis maps out the cumulative production of all the participants
against their cost of production on the vertical axis. By overlaying the current price of
the commodity onto the chart it can provide a quick indication of which producers are
currently profitable. As a rule of thumb some analysts will focus on those producers sitting beyond the 90th percentile responsible for 10% of all production. These are treated
as the marginal producers who in theory would withdraw supply from the market in
the event of prices being persistently below their costs.
In theory, marginal producer cash operating costs are a good guide to the floor for
prices during periods of oversupply. When prices fall to or just below marginal production costs, a portion of output will incur losses leading to the removal of supply from
the market. This cuts the available supply and prevents prices from falling further. However, marginal production costs are not a good guide to prices during periods of rapid
demand growth when supply is struggling to keep up with demand.
It is important to realise that these curves are not static:
‘Historically cost factors appear to have been highly dynamic, with the marginal
cost of production (and consensus long-term forecasts) tending to move up with spot
Cost of production
90th percentile
Production
FIGURE 5.3 Hypothetical marginal cost curve.
Base Metals
137
prices and then falling as the cycle turns down. Costs (like all market prices) are endogenous to the broader economic environment’ (Credit Suisse, 2013).
5.8.5
Premiums
Metal premiums are charged by a producer to a customer. They are designed to cover the
cost of shipping metal to a customer and include elements such as transportation, warehousing, financing, alloying, and marketing. They are market driven and negotiated
with the client as part of the terms of a commercial contract. They represent revenue
for a refiner and a cost to a consumer. The easiest way to understand the role of the premiums is to realise that the LME price is, in effect, a ‘global price’ that does not always
capture regional differences. Suppose that a US consumer of aluminium is negotiating
a bilateral supply contract with a producer. When it comes to establishing the price to
be paid, this could be made up of several components:
▪ The LME official price for a specific date or an average of LME prices over a specific
period.
▪ A regional premium which is intended to reflect the availability of the metal in a
particular geographic region and the costs associated with delivery.
▪ Premiums which may reflect the costs of producing a particular shape, quality, or
alloy.
▪ Any foreign exchange rate considerations (e.g. a EUR-based consumer who is paying in USD).
▪ The part of the supply chain to which the contract relates (e.g. ores and concentrates
may trade at a discount to the LME price, while a finished product with a higher
specification than the LME might attract a premium).
The sum of these different components would represent the ‘all in’ price paid for
the metal.
During the preparation of this manuscript (mid-2018), the USA announced plans
for tariffs on steel and aluminium imports. One of the consequences was that since
the US is a net importer of aluminium, the US Midwest Aluminium futures premium
registered a significant increase in anticipation of higher costs either from increased
demand for US-produced metal or imported metal subject to tariffs.
Premiums for aluminium in the US market came under regulatory scrutiny in 2014
when the commodity-related activities of several market participants were reviewed10 .
During the investigation it was argued that warehouse queues were artificially managed to increase the amount of time it took to remove metal from a particular location
increasing the warehouse owners’ rental income. The report noted that the increase
in the number of days taken to ‘load out’ the metal was associated with an increase in
the regional premium. The implication being that this extra rental cost was one factor
increasing the regional premium.
10
US Senate (2014)
138
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
One of the more recent innovations from the LME has been a contract that allows
participants to hedge the size of the premium for aluminium. The contract specification
for the US aluminium premium contract is shown in Table 5.4.
TABLE 5.4 Aluminium premium futures contract.
Feature
Description
Contract name
Region
Underlying metal
LME Aluminium US premium
Midwest and South US regions
High grade primary aluminium premium warrant in the designated
region.
25 tonnes
Third Wednesday of each maturity month
Monthly out to 15 months
US dollars per tonne
Physical delivery Seller provides: LME aluminium premium warrant
in the designated region. Buyer provides: Any LME warrant, plus
the premium cash as agreed at the time of the contract formation,
less the “Free on Truck” charge at the warehouse where the LME
premium warrant is delivered.
Lot size
Prompt dates
Maturity months
Price quotation
Settlement type
Source: Modified from LME, London Metal Exchange- Aluminium, Future Contract Specifiation, 2020.
The aim of the contracts is to allow participants to hedge the ‘all in’ cost of buying
physical aluminium in a non-queued LME ‘premium warehouse’. The nature of the
transaction is considered in Figure 5.4.
Figure 5.4 illustrates the use of the contract in conjunction with a standard LME
contract. The standard contract is shown on the left-hand side of the diagram and illustrates the purchase of the metal where the buyer receives a standard LME warrant and
Standard LME contract
LME premium contract
Buyer
Buyer
Standard
metal
warrant
Cash
Standard
metal
warrant
+ premium
cash
Seller
FIGURE 5.4 Hedging a base metals transaction.
Premium
warrant
Seller
139
Base Metals
delivers the contracted price. The diagram on the right-hand side illustrates the purchase of a premium contract. From the buyer’s perspective their net position is that
they will pay the agreed cash amount plus a supplemental cash amount for the premium
contract in exchange for receiving a premium warrant in the designated region.
If a commercial contract does reference the LME price, then it is possible to
hedge this component using the exchange’s risk management products. Although,
the exchange offers a risk management contract to hedge the premium components,
it will reflect general demand and supply dynamics and may not capture the various
subtleties of an individual contract. As a result, there could be an element that is
unhedgeable, which represents the position’s basis risk.
5.9
HEDGES FOR ALUMINIUM CONSUMERS IN THE
AUTOMOTIVE SECTOR
The commodities primarily hedged by the automotive sector are:
▪ Aluminium (body panels, engine blocks).
▪ Copper (wiring).
▪ Zinc (stainless steel coating).
▪ Lead (batteries).
▪ Platinum and Palladium (catalytic converters).
▪ Plastics (cabin interiors).
▪ Steel (body panels).
Of these commodities, aluminium is the most significant in terms of volume
because the physical characteristics of the metal are attractive. It offers a good strength
to weight ratio and since this reduces the weight of the vehicle, there is an added
environmental benefit in that the level of emissions are lower.
One of the ways in which the aluminium content could be hedged is by using the
NASAAC (North American Special Aluminium Alloy Contract) future on the LME.
This futures contract was set up by the exchange to meet the specific hedge requirements of the US automotive sector. The contract specification is ‘engine block alloy’
and delivery locations are near the end users. It was introduced due to concerns that
the price movements of the primary aluminium alloy contract was not sufficiently correlated with the actual metal used by the industry. However, to date, the contract has
experienced mixed success mainly due to poor liquidity. Although the car producers
would be natural buyers of the contract, there are limited natural sellers. The sellers
tend to be scrap merchants who are of poor credit quality and are therefore constrained
as to the tenor to which they can sell forward.
5.9.1
Forward purchase
For purposes of illustration we will assume the following prices for aluminium (USD
per metric tonne).
Cash:
Three months:
USD 2,400
USD 2,500
140
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The use of averaging to settle commodity market contracts is a common practice
and has several benefits. Take for example a consumer of metal who decides to link
the cost of buying the physical metal to some form of average. By using the average
price, the cost of the metal will not be based on an extreme level, as a sharp upwards
movement prior to the end of the period will be averaged out. However, by the same
token they cannot take advantage of a sharp downward movement, which would reduce
their cost.
The simplest (and often the most popular) method of hedging a purchase is to buy
the metal for forward delivery. In this example we will separate the delivery of the metal
from the hedge. That is, the consumer will buy the required tonnage of the metal from
their normal supplier and enter a cash-settled hedge transaction with a bank. Since the
manufacturer will take delivery of the metal from his regular supplier in the future he
is said to be ‘short the metal for forward delivery’. That is, a rise in the price of the metal
will incur extra costs while a fall in the price will save money. For the purposes of this
example, it is assumed that the commercial contract will be in the form of a ‘floating
forward’. This means the consumer agrees to buy the metal for delivery at a fixed future
date, but the settlement price will only be determined at the delivery date and will be
based on some average price. There are many ways in which this could be structured;
for example, a three-month forward deal with the settlement price based on the average
of the LME cash price over the month prior to settlement. In addition to this physical
contract the consumer agrees a cash-settled, three-month floating price forward with
its bank at a price of USD 2,500/metric tonne. This contract will also settle against the
average of the LME cash price over the month prior to settlement.
The forward price hedge with the bank is cash settled, as it is not intended as a
mechanism for obtaining physical supplies of the metal. In this case it is assumed that
the cash settlement amount on the forward is simply the difference between the average
LME cash price in the month prior to maturity and the forward price agreed at the outset
of the hedging transaction. Somewhat confusingly, the market will refer to this type of
deal as a ‘swap’ transaction as it is, in effect, a single period contract for difference.
However, it should not be confused with a multi-period swap or a swap transaction
where two entities execute a simultaneous purchase and sale for different maturities
against cash.
At the end of the three-month period we will assume that the appropriate monthly
average LME cash settlement price is USD 2,400. Under the terms of the physical contract, the consumer takes delivery of the required amount of metal at this price from his
supplier and settles the forward contract in cash with the bank.
In this case the settlement price is below the forward and so the manufacturer must
pay USD 100.00 (USD 2,400 – USD 2,500) per tonne to the bank. However, from the
manufacturer’s perspective, their net expense is the cost of buying the underlying metal
from their normal supplier (USD 2,400) plus the cash payment to settle the forward
contract with the bank (USD 100) – a total cost of USD 2,500, which was equal to the
fixed price agreed under the terms of the hedge with the bank.
Had the settlement price been USD 2,600 then the manufacturer would pay this
amount to procure their physical supply but would receive USD 100.00 from the bank
in settlement of the hedge. Again, the net price achieved for the purchase of the metal
would be USD 2,500.
Base Metals
5.9.2
141
Carry trades in the base metal market
One of the issues faced by end users of derivatives relates to the timing of the hedge. Let
us assume that the aluminium supplier tells the manufacturer that there will be a delay
in the shipment of the metal. This means that the hedging transaction also needs to be
delayed avoiding a mismatch in timing. Closing out the original trade can restructure
the hedge with an equal and opposite transaction and simultaneously execute a second transaction that re-establishes the original exposure, but at a different time in the
future.
Recall that the original hedge required the consumer to buy the metal for
three-month forward delivery at USD 2,500. After one month the consumer decides
that he wants the forward delivery date to be moved back by a month. The original
three-month exposure now has a residual exposure of two months, and so the consumer sells an offsetting two-month forward and simultaneously buys the metal for
three-month delivery. This simultaneous sale and purchase of futures for different
maturities is referred to generically as a ‘carry’ trade. When the consumer sells the
future for a shorter-dated maturity and buys a future for a longer-dated maturity, it
it is classified as a ‘lend’. This is because when the combination of forwards used to
switch the delivery are analysed from an economic perspective, the metal is being
delivered out for one settlement date but being repurchased for delivery at a future
date. This is similar in structure to a gold swap analysed in Chapter 4. Readers familiar
with the foreign exchange markets would recognise this type of transaction as an
FX swap.
An example will illustrate the concept. Let us assume that after one month the
market is still in contango and the prevailing two- and three-month forward prices are
USD 2,600 and USD 2,650, respectively. From the consumer’s perspective the original
long forward at USD 2,500 is closed out by selling a two-month forward at USD 2,600 to
yield a profit at the prompt date of USD 100.00. Simultaneously, the consumer initiates a
new long position for delivery in three months at the now prevailing price of USD 2,650,
which is higher than the original forward price of USD 2,500. However, the hedge cost
will be reduced by the USD 100.00 profit received on close out of the original exposure.
This would give a new forward purchase cost of USD 2,550 (USD 2,650 – USD 100.00) if
one ignores the time value of money. If the profit of USD 100.00 made from closing out
the original futures position could be put on deposit to earn interest between months
two and three, the final purchase cost would be further reduced.
The opposite of a lend is a borrow; a combination strategy where the metal is bought
for near-dated future delivery and sold for a far-dated future delivery.
5.9.3
Vanilla option strategies
One of the drawbacks for a consumer buying metal on a forward basis is that if
the price of the underlying metal declines, they would be locked into paying a
settlement price above the current market. The alternative is an option-based strategy.
All the following strategies are priced using the following assumptions
142
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Cash price
Three-month forward
Option style
Implied volatility
Maturity
USD 2,400
USD 2,500
European, cash-settled
20% p.a. (no skew)
Three months (92 days)
The terms of the options are designed to mirror those of the underlying contract to
buy the physical metal.
5.9.3.1
Synthetic long put
In the previous section the manufacturer was short metal for forward delivery by virtue
of his physical contract with the supplier. To hedge this exposure the manufacturer
could buy a call option with the strike set a level that suits their risk appetite. In many
cases an out-of-the-money (OTM) strike is selected to reduce the premium cost. We will
assume that the manufacturer decides to execute the call with a strike of USD 2,600,
which makes the option OTM with respect to the three-month forward price of USD
2,500. The premium for this option is USD 60.00. Figure 5.5 shows the ‘at expiry’ position for the manufacturer.
The vertical axis is defined as the cost of buying the metal while the horizontal
axis shows the current price of the metal. The definition of the vertical axis may seem
somewhat counterintuitive, but this approach allows for the diagram to incorporate the
individual option payoffs using the traditional ‘hockey stick’ orientation.
Decreasing
cost
Underlying
position
Long call option
Increasing
cost
FIGURE 5.5 Synthetic long put.
3,027.4000
2,976.2500
2,925.1000
2,873.9500
2,771.6500
2,822.8000
2,720.5000
2,669.3500
2,618.2000
2,567.0500
2,515.9000
2,464.7500
2,413.6000
2,362.4500
2,311.3000
2,260.1500
2,209.0000
2,157.8500
2,106.7000
2,055.5500
Net position
U/L Price
143
Base Metals
Notice how the net position resembles a long put option. Recall the shorthand version of put-call parity (see Chapter 2):
+C − P = +F
Where:
+C = Long call option
−P = Short put option
+F = Long position in the underlying asset
Although strictly speaking this condition only holds if the strike, maturities, and
notional amounts are the same, applying the principles intuitively we can rearrange
the formula to arrive at:
+C − F = +P
That is, the combination of buying a call option and a short forward position will
be equivalent to a long put.
The maximum cost to the manufacturer occurs when the market price rallies.
Assume that the settlement price of the metal at the option’s expiry is USD 2,700. The
manufacturer:
▪ Buys the physical metal from their normal source and pays the settlement price of
USD 2,700.
▪ Exercises its call option and cash settles the contract. The manufacturer receives
USD 100.00 per tonne (USD 2,700 − the strike of USD 2,600).
This returns a net amount of USD 2,600, a sum equal to the strike of the option. To
calculate the next cost the premium paid on the option must be added to this amount.
Ignoring time value of money considerations (the premium was paid at the start of the
contract and the net cost is calculated at the expiry of the option), the maximum cost of
the metal is USD 2,660 (USD 2,700 − USD 100.00 + USD 60.00).
A common misconception about this strategy is that it would be appropriate if the
manufacturer believed that the underlying price was about to rise. If this were the case,
a more effective strategy would be to buy the metal forward at the prevailing price of
USD 2,500, as it does not incur an option premium.
The purchase of the call option would be appropriate if the manufacturer believed
the price was going to fall but could not afford for his view to be wrong. If the price of the
metal falls, he forgoes the option and buys the metal in the underlying market at a price
lower than the strike, although the premium would represent a sunk cost. If the metal
rises in value against his expectations, the call option offers protection by ensuring a
maximum purchase price.
5.9.4
Short option positions
As in the previous examples we will assume that the consumer is short the metal for
forward delivery.
144
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
5.9.4.1
Selling out-of-the-money (OTM) puts
The manufacturer could buy the physical metal for delivery in three months’ time and
sell an OTM put option with the same maturity (see Figure 5.6). Using the principles of
put-call parity this net position resembles a short call. Since,
+C − P = +F
So,
−F − P = −C
Suppose that the put option is sold at a strike of USD 2,450 generating a premium of
USD 76.00 per tonne. This premium can then be used to subsidise the eventual purchase
price. If the price of the metal falls below the strike price, the company would have to
compensate the bank for the difference. This would end up increasing the cost of the
forward purchase. To illustrate, let us assume the price of the metal falls to USD 2,300
per tonne. The manufacturer:
▪ Buys one tonne of the physical metal for USD 2,300 from his usual supplier under
the terms their physical contract.
▪ Cash-settles the put option, paying the difference between the strike (USD 2,450)
and the spot price (USD 2,300) to the bank (i.e. USD 150.00).
▪ Retains the premium on the option (USD 76.00).
Decreasing
cost
Underlying
position
2,938.4000
2,888.7500
2,839.1000
2,789.4500
2,739.8000
2,690.1500
2,640.5000
2,590.8500
2,541.2000
2,491.5500
2,441.9000
2,392.2500
2,342.6000
2,292.9500
2,243.3000
2,193.6500
2,144.0000
2,094.3500
2,044.7000
1,995.0500
Short put option
U/L Price
Net position
Increasing
cost
FIGURE 5.6 Short forward position combined with sale of OTM put.
145
Base Metals
Ignoring the timing of the premium payment, the total cost of buying one tonne of
the metal is: USD 2,374 (USD 2,300 + USD 150.00 − USD 76.00)
However, if the price of the metal were to rise to an average price of USD 2,700 per
tonne the cost of buying the metal would be:
▪ The cost of purchasing the physical metal at a price of USD 2,700
▪ Less the USD 76.00 premium received on the option, which is not exercised
This results in a cost of USD 2,624 per tonne, so the result is that as the underlying
price rises, the net cost of buying the metal is subsidized by the option premium.
This strategy will reduce the forward cost of the metal if the manufacturer expects
the price of the metal to rise. If the settlement price for the metal was below the strike,
then on a net basis the consumer will pay a fixed price.
5.9.4.2
Selling OTM calls
In this strategy the manufacturer would buy the metal for delivery on a forward basis
and sell a cash-settled call option (see Figure 5.7). The strike of the call option would
be set at a price level that the manufacturer does not expect to trade. If the price of the
metal falls the options are not exercised and the premium income reduces the purchase
cost. A similar situation arises if the underlying market rises but does not hit the strike.
If the market trades above the strike, the calls are cash-settled incurring a cost. Notice
that the position is relatively risky compared to the previous examples. If the price of the
metal falls, the consumer will buy the metal at a lower price and will retain the option
premium lowering the net purchase price. A rise in the underlying price beyond the
Decreasing
cost
Short call option
Increasing
cost
FIGURE 5.7 Short forward position combined with sale of OTM call.
3,027.4000
2,976.2500
2,925.1000
2,873.9500
2,771.6500
2,822.8000
2,720.5000
2,669.3500
2,567.0500
2,618.2000
2,515.9000
2,413.6000
2,464.7500
2,362.4500
2,311.3000
2,260.1500
2,209.0000
2,157.8500
2,106.7000
2,055.5500
Underlying
position
U/L Price
Net position
146
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
strike of the sold option results in them having to pay increasing amounts for both the
physical metal and the short option position.
5.9.4.3
Selling swaptions
Suppose a miner is running a host operation, which is currently not working at
full capacity, possibly due to low market prices. The miner would be willing to run
at 100% capacity if the price of metal were 15% higher than its current value. This
suggests that there is an element of optionality within the operation; the ability but
not the obligation to produce. The miner sells a cash-settled call option on the metal.
If prices rise and the option is exercised, the buyer will receive the difference between
the spot and strike price. To offset this cash payment on the derivative the miner will
have to increase production, but this is now down at levels that are economic.
5.9.5
Combination option strategies
Arguably the main motivation behind combination strategies is to offer the client a
hedging solution at a lower or no premium cost.
5.9.5.1
‘Min-max’
The min-max option strategy is probably the most common combination strategy seen
across all asset classes. As with many option strategies it has different names depending
on the underlying asset class: zero-premium collar, range forward, and cylinder are just
some of the phrases used to describe the strategy.
One way of constructing the strategy is:
▪ Purchase the physical metal for delivery in three months’ time.
▪ Buy out of the money calls at a strike of USD 2,600 to give upside price protection.
▪ Sell OTM puts with the strike set at such a level that the net premium is zero.
Based on the prevailing market parameters, the strike of the sold put would be set at
USD 2,410. At first glance, it is not necessarily obvious what the option solution offers.
However, recall that both short calls and long puts are selling strategies. The strategy
fixes the minimum sale price (via strike of the long put) and the maximum sale price
(via the short call). If the settlement price applicable to both the metal and option rises
to USD 3,000, then the cost of purchasing the metal is reduced by virtue of the long call
which pays out the difference between the strike and the final price (i.e. USD 400.00).
The put is not exercised leaving the consumer with a net purchase cost of USD 2,600.
See Figure 5.8.
By the same logic, the impact of the short put position means that if the price of the
metal were to fall, the consumer would not be able to benefit from price falls beneath
the strike of this component of the transaction. So if the settlement price for the option
and the metal was USD 2,000, the put is exercised against the manufacturer who pays
USD 410, resulting in a net cost of USD 2,410, which is an amount equal to the strike of
the sold put.
In between the two strikes neither option is exercised, and the consumer simply
buys the metal in the market at the prevailing price.
147
Base Metals
Decreasing
cost
Long call option
Underlying
position
2,972.3417
2,922.0917
2,871.8417
2,821.5917
2,771.3417
2,721.0917
2,670.8417
2,620.5917
2,570.3417
2,520.0917
2,469.8417
2,419.5917
2,369.3417
2,319.0917
2,268.8417
2,218.5917
2,168.3417
2,118.0917
2,067.8417
2,017.5917
Short put option
Net position
Increasing
cost
FIGURE 5.8 ‘Min-max’ solution.
5.9.5.2
Ratio min-max
In this structure illustrated in Figure 5.9, the strike on the purchased call is kept at USD
2,600. The transaction is still structured to have zero premium, but instead of selling
options in equal proportions to the calls, a smaller notional of puts are executed. In
this example puts equal to 75% of the underlying exposure are sold, but to ensure the
structure is zero premium the strike of the put options has to be moved closer to the
current forward price to increase the premium. In this instance it would result in a
strike of USD 2,458 on the sold put.
As a result, there is no longer a minimum price level at which the hedger will buy
the metal. However, they will not get 100% of the benefit of a falling price as the sold
put option represents a liability. Since the notional amount on the sold put is 75% of
the call option, a one-unit fall in the price below the strike of the put will result in
the hedger benefiting by 25% of this change. That is, they can only benefit from the
unhedged portion of the transaction.
At first sight this strategy is superior to the simple min-max as it does not impose a
floor on the purchase of the metal in a falling market. However, like all option strategies
it is not possible to obtain something for nothing and there is indeed a trade-off. With
the simple min-max solution, the hedger will benefit fully from any decline in price
up until the strike of the put. For the ratio min-max solution, although the hedger will
benefit in a falling price environment below a relatively higher threshold level, they
will only benefit from 25% of the decline, albeit with no floor.
5.9.5.3
‘Three way’
In the ‘three way’ structure (Figure 5.10), the trade consists of a bull spread (constructed
using calls) financed by the sale of a put. The structure is zero premium.
148
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Decreasing
cost
Underlying
position
Long call option
2,991.1575
2,940.6075
2,890.0575
2,839.5075
2,788.9575
2,738.4075
2,687.8575
2,637.3075
2,586.7575
2,536.2075
2,485.6575
2,435.1075
2,384.5575
2,334.0075
2,283.4575
2,232.9075
2,182.3575
2,131.8075
2,081.2575
2,030.7075
Short put option
Net position
Increasing
cost
FIGURE 5.9 Ratio ‘min-max’ solution.
A bull spread is created by the purchase of a low strike call and the sale of a high
strike call11 . Since the purchased call is in-the-money and the sold call out-of-themoney, the combined position has an associated cost. This is financed by the sale of
a put option with the strike set at a level that results in a zero premium. Using the
previous data, one possible permutation is:
▪ Short a put option at USD 2,376,
▪ Long a call option at USD 2,450,
▪ Short a call option at USD 2,550.
The final cost to the consumer will vary according to the settlement price at
maturity:
▪ If the price is below USD 2,376, the cost of buying the metal will also be USD 2,376
by virtue of the sold put.
▪ If the market price is between the strike of the sold put (USD 2,376) and the purchased call (USD 2,450), the consumer pays the prevailing market price as neither
option is exercised.
▪ For market prices between USD 2,450 and USD 2,550, the consumer will pay a strike
equal to the purchased call as neither of the sold options is exercised.
▪ Above the strike of the sold call (USD 2,550), the consumer will pay the prevailing
price less USD 100.00. This USD 100.00 cost reduction is attributable to the bull
11
Bull spreads are sometimes referred to as ‘call spreads’.
149
Base Metals
Decreasing
cost
Underlying
position
Long call option
2,746.3040
2,716.6040
2,686.9040
2,657.2040
2,627.5040
2,597.8040
2,568.1040
2,538.4040
2,508.7040
2,479.0040
2,449.3040
2,419.6040
2,389.9040
2,360.2040
2,330.5040
2,300.8040
2,271.1040
2,241.4040
2,211.7040
2,182.0040
Short call option
Net position
Short put option
Increasing
cost
FIGURE 5.10 ‘Three way’ structure.
spread element of the strategy and is a reflection that prices above USD 2,550, the
purchase of a call at USD 2,450 and the sale of a call at USD 2,550, would result in
both options being exercised. Since both options are exercised, the consumer will
end up with a profit of USD 100.00, which is the difference between the two strikes.
5.9.6
Structured option solutions
The range of option-based strategies widens considerably if one were to include structured solutions that are constructed using exotic options such as barriers. Three possible
strategies are highlighted here with an explanation as to how they are structured. As
with the previous examples, it is assumed that the consumer has an agreement to buy
the metal at a future date based on the price that prevails at the time. As such they are
‘short’ the market for forward delivery.
5.9.6.1
Knock-out forward
This is a standard forward contract that automatically terminates if the cash price trades
at, or beyond a predetermined knock out price before expiry. To compensate the investor
for the potential early termination of the contract, the forward price is set at a more
favourable level than the prevailing market price. The knock-out forward is generally
structured for zero upfront cost.
Market prices
Three-month forward price
USD 2,500
150
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Contract prices
Contract forward price
Knock-out price
USD 2,450
USD 2,932
In this transaction the consumer will buy aluminium at a price of USD 2,450 at
expiry of the contract as long as the knock-out price of USD 2,932 does not trade during
the life of the transaction. If the knock-out price trades, the forward contract is terminated and the consumer is left unhedged at what would now be a disadvantageous price
compared to the original three-month forward. See Figure 5.11.
The construction of this strategy is based on the principles of put-call parity using
barrier options. The contract forward price is created by the purchase of a knock-out
call and the sale of a knock out put with the same strike and barrier.
+C − P = +F
If the two options continue to be active, they combine to give the favourable synthetic forward position. If the knock-out price trades, resulting in the termination of
both options, then the consumer is left unhedged in an unfavourable set of market circumstances, i.e. higher prices.
5.9.6.2
Forward plus
In a forward plus contract the client obtains protection against a steep rise in the
underlying price by virtue of a zero-premium call option that is struck slightly
Decreasing
cost
3,308.1667
3,233.9167
3,159.6667
3,085.4167
3,011.1667
2,936.9167
Short knock out
put option
2,862.6667
2,788.4167
2,714.1667
2,639.9167
2,565.6667
2,491.4167
2,417.1667
Long knock out
call option
2,342.9167
2,268.6667
2,194.4167
2,120.1667
2,045.9167
1,971.6667
1,897.4167
Underlying
position
Net position
Increasing
cost
FIGURE 5.11 Knock-out forward.
151
Base Metals
out-of-the-money. However, the contract contains a trigger level set below the current
market price, which if activated, will alter the payoff of the structure. If the cash price
trades above this trigger rate during the life of the transaction, the client will retain the
right to buy the underlying at the call strike. However, if the cash rate trades below the
trigger price, this right becomes an obligation with the client being required to take
delivery of the underlying at the call strike price.
To illustrate the concept let us assume the following values are observed in the
market:
Market price
Three-month forward price
USD 2,500
Strike price
Trigger level
USD 2,600
USD 2,139
Contract price
If the cash price rises during the life of the transaction and the trigger level does
not trade, the consumer will buy aluminium at the strike price of USD 2,600. This is
less advantageous than the initial forward rate, but this optionality is obtained at zero
premium. If the cash price falls below the strike but remains above the trigger level, the
consumer can walk away from the call and buy the metal at the lower market price.
However, if the spot price falls to the trigger level of USD 2,139, the option to buy at
the strike of USD 2,600 now becomes an obligation. This means that the consumer will
never pay more than USD 2,600 for the metal but can only benefit from a fall in the
price to USD 2,139. The payoff is illustrated in Figure 5.12.
This transaction is structured using the following components:
▪ Consumer buys a European-style call option with a strike of USD 2,600 at a cost of
USD 60.00/tonne
▪ Consumer sells a knock in put option with a strike of USD 2,600 and a trigger level
of USD 2,139 at a cost of USD 60.00/tonne
If spot does not trade at or below the trigger level, the consumer is left holding
the European call. However, if the trigger level trades and the short put is activated, the
combination of options creates a synthetic long-buying position, again through the principles of put-call parity (i.e + C − P = + F). The synthetic forward requires the consumer
to buy the metal at the strike price of the two options.
5.9.6.3
Basket options
Since an automotive producer may have a requirement to buy several different metals,
an option-based risk management strategy based on individual call options could prove
to be very expensive.
152
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Decreasing
cost
Underlying
position
Long call option
3,236.1667
3,158.9167
3,081.6667
3,044.4167
2,927.1667
2,849.9167
2,772.6667
2,695.4167
2,618.1667
2,540.9167
2,463.6667
2,386.4167
2,309.1667
2,231.9167
2,154.6667
2,077.4167
2,000.1667
1,922.9167
1,845.6667
1,768.4167
Short knock-in
put option
Net position
Increasing
cost
FIGURE 5.12 Forward plus structure.
An alternative structure that reduces the cost of protection is an option on a basket
of metals. In the following example we will consider an equally weighted, two metal
basket option based on copper and aluminium.
The payoff on this basket option at maturity is:
Basket call option: Max [0, (Q1 S1 + Q2 S2 ) − X]
Basket put option: Max [0, X − (Q1 S1 + Q2 S2 )]
Where:
Q = weight of metal within the basket
S = price of the metal at expiry
X = strike price
The strike price of the option, which is set at the start of the deal, is a weighted strike
price. That is, the strike chosen for each individual metal is weighted by the proportion
it contributes to the basket.
Suppose that the basket strike is based on the current observed three-month futures
price for both metals with the respective values being USD 2,500 for aluminium and
USD 7,000 for copper. The strike rate for an equally weighted basket would be USD
4,750 (50% x USD 2,500 + 50% x USD 7,000).
If a consumer were to buy a call option on this particular basket and the final prices
at maturity were USD 2,600 and USD 7,100 for aluminium and copper respectively,
153
Base Metals
ignoring premium costs, the value of the basket would be USD 4,850 (50% x USD 2,600
+ 50% x USD 7,100). The payout on the option would be USD 100.00 (USD 4,850 − USD
4,750).
In the payoff example illustrated above, note that the basket option paid out a total
of USD 100.00 for the two metals. Individual call options on each metal would have paid
out USD 200.00. Since the payout on the basket is lower than the two individual calls,
then the premium cost will be lower.
The premium cost of a basket option is based on the premise that the constituent
prices will never all rise and fall in tandem. That is, the correlations between the underlying metals in the basket work as a form of a natural hedge.
When pricing a basket option, the key difference in the product’s valuation is
the implied volatility input, which incorporates the price correlation between the
constituent components. Price correlation describes the tendency of prices to move in
the same direction and is expressed along a scale from +1 to −1. Positive correlation
implies that both asset prices will tend to move in the same direction, whilst assets that
have negative price correlation will tend to move in opposite directions.
The form of the implied volatility input has been adapted from portfolio theory (see
Schofield, 2017) and is expressed as follows for a two-asset basket:
√
𝜎basket = (w2 x1 𝜎x21 ) + (w2 x2 𝜎x22 ) + 2 × (wx1 wx2 𝜌x1 x2 𝜎x1 𝜎x2 )
where:
𝜎x21 = Variance of asset 1
𝜎x22 = Variance of asset 2
𝜌x1 x2 = Correlation between asset 1 and asset 2
𝜎x1 = Volatility of asset 1
𝜎x2 = Volatility of asset 2
wx1 = Proportion of asset 1
wx2 = Proportion of asset 2
If we assume implied volatilities of 20% for aluminium and 25% for copper, the
composite basket volatility for a range of different correlation values and the associated
premium for a three-month option with a strike of USD 4,750 are:
Price correlation
−1.0
−0.5
0.0
+0.5
+1.0
Basket volatility
Option premium
2.50%
11.46%
16.01%
19.53%
22.50%
USD 125.00
USD 155.00
USD 181.00
USD 204.00
USD 224.00
154
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
All other things being equal if correlation increases, the volatility of the option will
increase making the option more expensive. Intuitively, if the two assets are positively
correlated, then their prices are more likely to move in the same direction, increasing
the magnitude of any expected payment to be made by the seller of a call option. Since
the premium on an option can be thought of as the present value of the expected payout,
this makes the option more expensive. The option will be less expensive where the asset
prices are negatively correlated, as price movements will tend to be offsetting, reducing
the expected payout.
The cost of the basket option can now be compared to the cost of two individual
at-the-money forward three-month call options.
Aluminium
Strike
Implied volatility
Premium
USD 2,500
20%
USD 99.00
Strike
Implied volatility
Premium
USD 7,000
25%
USD 348.00
Copper
The total cost of purchasing two individual call options is therefore USD 447.00,
compared to the maximum cost of the basket option of USD 224.00. Note that there is a
relationship between the two individual call options’ assets and the basket option priced
with a correlation of +1. At this level of correlation, the implied volatility of the basket
option is 22.50% equal to the weighted average of the two volatilities of the individual
call options. Also, the premium on the basket option is equal to the weighted average
of the premiums on the individual options. However, there is a small difference due to
the fact that the individual call options were priced with a closed-form model, whereas
the basket option was priced with a binomial model, whose output is sensitive to the
number of steps used.
However, note that the premium on the basket option is always lower than the
individual options because of correlation. Even though anecdotally, it may be perceived
that prices of metals may move in the same direction, it is unlikely that they will be
perfectly positively correlated.
5.9.7
Foreign currency exposures
5.9.7.1
Vanilla FX hedge
Suppose that a consumer has entered a three-month fixed forward contract with a bank
to sell 1 tonne of copper with the price expressed in EUR. At the time of the transaction
155
Base Metals
the three-month copper price is USD 7,000. Since the trader must quote the sale in a foreign currency, the first step is to reprice the contract in EUR at the current three-month
forward FX rate. In this example, the applicable rate is assumed to be €1 = USD 1.20.
This will fix the cost of the metal to the client at EUR 5,833.
The transaction now leaves the bank exposed to movements in the price of the metal
as well as movements in the foreign exchange rate.
The hedging of the metal price risk by the bank could be achieved by either buying it
forward from another market participant or buying it now and holding it until maturity.
In this example, however, we are less concerned with the metal price risk but rather the
foreign exchange rate risk.
If the bank does hedge the metal price risk, the trader is due to receive EUR 5,833
in three months’ time from the client, but their own cost of procuring the metal will
be USD 7,000. If left unhedged the trader will be exposed to movements in the spot
exchange rate. For example, if left unhedged and the exchange rate moved to €1 =
USD 1.15 then in three months’ time, the bank would receive EUR 5,833 from the client
to settle the metal transaction which would then be sold in the FX markets for USD. At
this spot exchange rate, the amount of USD received by the bank would be USD 6,708,
which is lower than the amount required to settle the forward purchase of copper from
the metals market.
To illustrate how this could be hedged the different components of the structure
are shown in Figure 5.13.
Suppose that the following rates are noted in the FX market:
▪ Spot €∕$ = 1.1962
▪ 3 m EUR interest rates = 0.25%
Metals market
$7,000
Cu
€5,833
Forward
FX market
€5,833
Bank
Client
$7,000
Cu
€5,833
Lend €@ 0.25%
Spot FX market
Bank
$6,977
Money market
Borrow USD @ 1.50%
FIGURE 5.13 Hedging the FX exposure of a non-USD sale of metal.
156
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ 3 m USD interest rates = 1.50%
▪ 3 m forward €∕$ = 1.20
In the upper part of the diagram the trader executes three transactions:
1. Sells the client their copper at an agreed EUR price of €5,833∕tonne for delivery in
three months’ time.
2. Buys copper three month forward at USD 7,000/tonne.
3. Sells the EUR 5,833 proceeds of the client deal in the forward FX market to receive
the USD 7,000 equivalent, which can be used to settle the metals transaction.
However, conventions in the wholesale FX market mean that it is unlikely that the
bank will be able to hedge the FX exposure with a single forward transaction. Since this
deal effectively pushes the exposure to spot rates to the other FX counterparty, such
trades are executed with a simultaneous spot FX deal. This transaction is shown on the
left-hand side of the bottom part of the diagram. The combination of the two FX deals is
referred to as an FX swap. Although it is beyond the scope of this textbook to illustrate
this, the bank will have a residual risk to the ratio of interest rates between the two
currencies. The final part of the diagram on the bottom right-hand side of the figure
acknowledges that not only will the bank have to borrow the USD they are required to
deliver on the spot leg, but they will also need to lend the EUR received in return.
5.9.7.2
FX-linked commodity swap
This structure allows a non-USD market participant to hedge the FX component of its
commodity exposures. These types of swap are sometimes referred to as ‘compo swaps’.
A possible structure is outlined in Figure 5.14.
To illustrate the example, we will consider the perspective of a Euro-denominated
copper consumer, based on a per tonne basis who agrees to a fixed exchange of USD 1.00
= EUR 1.20.
Suppose that at the point of settlement on the swap, the appropriate settlement price
of copper is USD 6,500 and the spot exchange rate is USD 1.00 = EUR 1.15. It is assumed
that the client has a separate commercial contract to purchase the metal in USD.
The cash flows on the commercial metal contract are:
Sell EUR 5,652 and buy USD 6,500 to purchase one tonne of metal
(Fixed volume of commodity * floating USD commodity price) / Fixed EURUSD exchange rate
Non-USD client
Bank
(Fixed volume of commodity * floating USD commodity price) / Floating EURUSD exchange rate
FIGURE 5.14 Commodity-linked FX swap (I).
157
Base Metals
(Fixed volume of commodity * fixed USD commodity price) / Fixed EURUSD exchange rate
Non-USD client
Bank
(Fixed volume of commodity * floating USD commodity price) / Fixed EURUSD exchange rate
FIGURE 5.15 Commodity-linked FX swap (II).
(Fixed volume of commodity * fixed USD commodity price) / Fixed EURUSD exchange rate
Non-USD client
Bank
(Fixed volume of commodity * floating USD commodity price) / Floating EURUSD exchange rate
FIGURE 5.16 Commodity-linked FX swap (III).
The cash flow payable by the bank in the swap is:
One tonne ∗ USD 6,500∕1.15 = EUR 5,652
The cash flow payable by the client in the swap is:
One tonne ∗ USD 6,500∕1.20 = EUR 5,417
The net result of the swap and the commercial contract means that the client delivers USD 6,500 to its supplier to receive one tonne of metal at a cost of EUR 5,417, i.e.
the EUR cost of the copper is fixed at a level of USD 1.00 = EUR 1.20.
A similar swap can be offered to the client, which hedges the commodity risk but
not the FX risk, and is shown in Figure 5.15.
If the client is looking to hedge both the FX and the commodity risks, then it is
possible to combine both of the structures into a single swap to reduce the number of
cash flows, as illustrated in Figure 5.16.
5.10
SUMMARY
Base metals are one of the most mature commodity markets and, as such, the principles
of risk management are well established within the market.
Given the multitude of different types of metals that exist within this segment, the
issue of pricing is a central theme. As a result, the role of exchanges such as the London
Metals Exchange (LME) and the Shanghai Futures Exchange become pivotal in terms
of price transparency and discovery.
158
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The chapter covered most of the common derivative strategies available to participants along the supply chain. The distinct features of the LME in terms of trading and
settlement were discussed as well as some of the contracts that are unique to the market
(e.g. futures on market premiums). Following on from this section, examples of popular
vanilla and structured option solutions were presented. The chapter concluded by considering the nature of the foreign exchange rate risks faced by banks who are willing to
offer risk management solutions in non-USD currencies.
CHAPTER
6
Crude Oil
6.1
OVERVIEW OF ENERGY MARKETS
Within commodities, the phrase ‘energy markets’ is often loosely used to describe a
series of markets, which include:
▪ Coal
▪ Renewables
▪ Electricity
▪ Nuclear
▪ Natural gas
▪ Crude oil
Another popular term is ‘hydrocarbons’, which is a term that could be used to
describe an organic compound that consists of carbon and hydrogen. This would
include coal, crude oil, and natural gas.
According to the BP Statistical Review of World Energy 2020, (Figure 6.1) the most
consumed form of energy is crude oil (33%), followed by coal (27%), and natural gas
(24%). Although much publicity is often given to the increase in the use of renewable
energy, particularly as a way of reducing the effects of climate change, the overall global
consumption figure from this source remains low at 5%.
Since each of these energy sources will be measured in different ways, BP creates
an element of equivalence by converting them into a standard measure, referred to as
millions of tonnes oil equivalent.
6.2
THE VALUE OF CRUDE OIL
The term ‘crude oil’ does not really describe any specific type of oil, rather the generic
state of oils prior to their refinement. For example, the price reporting agency, Platts lists
approximately 130 traded crude oils, each with different chemical characteristics that
are attractive for different reasons. Thus, when extracted from the ground crude oil may
be a pale straw-coloured liquid or a thick tar-like substance. The value of crude oil therefore lies in what can be produced from the refining process. For example, crude oil that
159
160
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Hydro
7%
Renewables
5%
Nuclear
4%
Crude oil
33%
Coal
27%
Natural gas
24%
FIGURE 6.1 Relative primary energy consumption by fuel
type.
Source: BP Statistical Review of World Energy 2020. BP PLC
is classified as ‘light’, (i.e. composed of short chain hydrocarbons) is sought for its relatively higher yield of gasoline.
Sometimes reference is made to ‘conventional’ and ‘unconventional’ oil. Generally
speaking, conventional oil comes from sources that are considered to be relatively
straightforward to extract using standard methods. Unconventional resources may
include sources such as shale oil or tar sands. These are more complex to extract and as
a result, production may only be economically viable once prices reach a certain level.
However, although shale oil can be ‘light and sweet’ (see following discussion on crude
oil characteristics), tar sands are considered to be low value bitumen and will need to
be upgraded in order that it will have characteristics similar to conventional crude oil.
6.2.1
Basic chemistry of oil
Crude oil is made up of hydrocarbons which are molecules comprised of hydrogen and
carbon atoms. The carbon atoms may be joined in a chain-like sequence or in a ring
formation, with different numbers of hydrogen atoms attached. A certain crude oil may
actually comprise of several different individual crudes, which are recovered from wells
that are geographically close together and therefore share the same extraction facilities.
However, the quality of crude oil is not a constant and could change over time, hence
different wells may contribute a variable quality in the overall supply. The quality of the
161
Crude Oil
crude oil is established through an assay, which is a laboratory-based assessment of the
quality of the crude. The assay will also establish the various boiling point fractions and
the percentage of each refined product that a particular crude can yield.
Since crude oil is of limited use in itself, its value lies in what can be produced
from the refining process. As a result, it would be fair to say that demand for crude oil
is effectively refinery demand. In addition, the value of each refined product may be
influenced by unique or independent factors. Take for example one of the more popular refined products such as gasoline. The chemical specification may be influenced
by factors such as government regulation (perhaps to reduce the lead content), current
market prices, local and seasonal demand.
When assessing the value of crude oil, several important characteristics are
considered.
6.2.2
Density
Density is a measure of the number of molecules within a defined volume. A simple
analogy would be to compare a glass of water of a given volume with the same volume
of mercury. Although they occupy the same space, the mercury will weigh considerably
more than the water due to its greater density. From a crude oil perspective, density will
help in quantifying the volume to weight ratio.
The density is usually expressed as an API gravity value. This is a measure of the
weight of hydrocarbons according to a scale devised by the American Petroleum Institute. Crude oils that have a high API value are less dense (lighter) and tend to produce
greater volumes of higher value products (e.g. the so called ‘distillates’ such as gasoline
and kerosene) as part of the refining process. Heavy crude oils with a lower API are
more difficult to refine in that they require more processing stages and energy, which
adds to the cost of production. They also tend to produce lower yields of the higher value
lighter products.
As a point of reference, water has an API of 10, so a fluid with a lower API is heavier
and will sink in water. A fluid with an API greater than 10 is lighter and will float on
water.
6.2.3
Sulfur content
Sulfur is an undesirable property when it appears in large quantities. Generally, crude
oil will be classified as being “sweet” when the sulfur content is less than 0.5%, although
other percentage values are sometimes cited (e.g. 0.7%).‘Sour’ crude is where the content
of sulfur by weight is greater than 0.5%. Crudes with a high sulfur content require more
processing and a greater energy input to complete the refining process.
6.2.4
Acidity
Acidity is measured using a Total Acid Number (TAN). High acid crudes are those with
a TAN greater than 0.7. Acidic crudes can be corrosive to refinery equipment and will
require greater investment to process.
162
6.2.5
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Flow properties
Viscosity is a measure of the ability of the crude oil or refined product to flow or its
resistance to pouring and can be measured on several different scales at a range of temperatures. Intuitively, it can be thought of as the friction of a liquid. For example, if a golf
ball were dropped into a jar of water, the resistance would be less than that experienced
if the ball were dropped into a jar of mayonnaise.
The pour point measures the lowest temperature at which either crude oil or a particular refined product flows as a liquid under a set of given conditions. A high pour
point indicates that the product has to be heated for it to flow as a liquid. This will have
an impact on how the oil is stored or transported.
6.2.6
Other chemical properties
Paraffinic – literally ‘like paraffin’, suggests that the crude has a low viscosity and high
flammability. This indicates that the crude could be used in the final production of lubricants. (In the UK, paraffin is the name given to kerosene).
Naphthenic – a crude oil with naphthenic properties is one that has a high viscosity but is not highly flammable. This might make it suitable for the production of, for
example, bitumen.
Intermediate – this describes a crude oil, which has properties that sit between
paraffinic and naphthenic properties.
6.2.7
Examples of crude oil
Light, sweet, low TAN oils are easier to process and tend to trade at premiums to heavier,
higher-sulfur, more acidic crudes which require additional treatment. Examples of the
different chemical characteristics of the crude oil that illustrate the principles are:
▪ Light and sweet crude (with an API of between 33–45) are West Texas Intermediate
(US crude oil) and Brent (North West European crude oil). Since they also have
lower wax content and fewer long chain molecules, they tend to be less viscous and
as such they are easier to pump and transport. They are attractive as production
and refining costs are lower and when refined will produce an attractive yield of
high-value gasoline.
▪ Heavy and sour (an API of between 10–45) crude oils include Urals crude (Eastern
European crude oil), which is used as bitumen feedstock. Heavy crudes tend to have
a higher viscosity with a consistency that can resemble molasses to even being a
solid at room temperature. They can be difficult to pump out of the ground or send
through a pipeline and may contain impurities such as sulfur and metals.
In its 2018 energy review ENI argue that the most commonly found quality of crude
oil was classified as being ‘medium and sour’, which equated to an API of between 26–35
with a sulfur content greater than 1%. This accounted for 41% of all production, while
the ‘light and sweet’ category amounted to 19% (Figure 6.2).
163
Crude Oil
Heavy and medium sour
2%
Heavy and sour Ultra light
4%
9%
Light and sweet
19%
Heavy and sweet
3%
Light and medium sour
5%
Light and sour
3%
Medium and sour
41%
Medium and sweet
10%
Medium and medium sour
4%
FIGURE 6.2 Relative crude oil demand by quality.
Source: ENI
6.3
AN OVERVIEW OF THE PHYSICAL SUPPLY CHAIN
The main elements of the crude oil physical supply are:
▪ Upstream activities – this is the exploration, development, and production of crude
oil from either an off- or on-shore location.
▪ Transportation and storage – the crude oil must be moved by pipeline or by tanker
to a particular refining location. If the refinery is close to the extraction point, the
oil can be stored locally prior to delivery or can be loaded onto ships for onward
delivery.
▪ Refining – as crude by itself is of limited use, it must first be refined into a variety of
products such as gasoline.
▪ Retail and wholesale distribution – the refined products can then be sold and delivered to the specific location for a wholesale purchaser (e.g. a particular airport for
an airline buying jet fuel) or to a network of retail petrol/gas stations. This part of
the supply chain can sometimes be referred to as the ‘marketing’ function.
▪ End consumer – either a retail participant (e.g. gasoline) or a wholesale client.
The main market participants will either be international oil companies (IOCs),
which are privately owned companies such as BP or state-owned national oil companies
(NOCs) such as Saudi Aramco. Although the privately owned companies are amongst
164
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
the largest corporates in the world (e.g. ExxonMobil), they are dwarfed in size by the
NOCs. There are also ‘independents’ that are primarily active in individual segments of
the market such as exploration, production, and storage.
Vertical integration is common along the supply chain, but it is not necessarily the
preserve of the private oil company. For example, Saudi Aramco is vertically integrated
with interests in exploration, production, refining, marketing, and international
shipping.
Other relevant parties who trade in the market are financial institutions providing
risk management solutions to supply chain participants and the commodity trading
houses (e.g. Vitol and Mercuria). Commodity trading companies sell to industrial consumers obtaining their products either from third party sources or production assets
they own. They may also engage in more speculative ‘view driven’ trading strategies.
Crude oil and the refined products are stored at oil depots, which are also sometimes
referred to as tank farms, installations, and oil terminals. From here they are transported
to the end users or to other storage facilities. They may be located close to oil refineries,
marine terminals, or at the terminus of a pipeline.
Traditionally crude oil and the refined products are moved mostly by ocean going
tankers and pipelines. Initial construction costs on pipelines are higher than ships but
offer lower variable costs. Where there are alternatives then pipelines, which tend to be
privately owned, is the preferred method of transportation.
Tankers used for liquid fuels are classified according to their capacity using the
AFRA system (average freight rate assessment) and whether they will carry crude oil
(referred to as ‘black’ products) or refined products (‘white’ products). The following
classifications are most commonly used for the oil market and are expressed in terms of
their dead weight tonnage (dwt – the weight in metric tonnes of the cargo, stores, fuel,
and crew):
10,000–60,000
60,000–80,000
80,000–120,000
120,000–200,000
200,000–315,000
320,000–550,000
Product tanker
Panamax
Aframax
Suezmax
VLCC (Very Large Crude Carrier)
ULCC (Ultra Large Crude Carrier)
As indicated by the names, some of the tankers when fully laden can traverse two
key shipping routes, the Suez and Panama canals. In terms of barrels of oil, a VLCC
can carry about two million barrels, while the ULCCs are able to cope with twice that
volume. The Product and Panamax tankers are mostly used for the transport of refined
fuels.
Where there is a suitable network of rivers or inland waterways, oil products can
be transported across large distances from the coastal refineries to supply domestic
markets. Such a system exists in North West Europe utilising the main river arteries
and their tributaries to supply local depots. The Rhine, the Seine, the Maas, and the
Danube all have significant oil barge traffic supplying central Europe. Barges move
much smaller quantities than tankers, typically between 1,000–3,000 metric tonnes.
165
Crude Oil
6.4
REFINING CRUDE OIL
6.4.1
What is refining?
From a chemical perspective no two crude oils are the same. They are a mixture of hundreds of different types of hydrocarbons (e.g. naphthenes and aromatics) with carbon
chains of different lengths, which are separated through the refining process.
6.4.2
What does a refinery produce?
Table 6.1 illustrates the main refined products with some typical uses.
6.4.3
Product yields
Refiners talk of a ‘product yield’ or a ‘product slate’, which describes the different proportions of refined products obtained from refining one barrel of crude oil. Yields may
vary slightly over the year if the refiners have the flexibility to respond to changing
market demand.
On average, a 42-(US) gallon barrel of crude oil will yield about 45 gallons of refined
products as the refining process generates a ‘refinery gain’. A tonne of heavy crude oil
will occupy a smaller space than a tonne of lighter crude. When the heavy crude is
therefore refined into a series of ‘lighter’ refined products, they will occupy a larger
TABLE 6.1 Refined products and their applications.
Gases and gas liquids – the most significant products in this category are methane, ethane, and
propane. Methane is commonly referred to as natural gas and can be used for heating. Ethane
is a feedstock for the petrochemical industry that can be used in the production of plastics.
Propane may be used for cooking and heating purposes.
Naphtha – is most commonly used as a petrochemical feedstock. It can also be used as a solvent
in dry cleaning fluids and in paint solvents. It is also an intermediate product that could be
further processed to make gasoline.
Gasolines – the main use of gasoline is to fuel motor vehicles. The gasoline will need to be of a
certain specification, the most widely known being the octane number, which is frequently
displayed on the forecourts of petrol stations. There are two ways of measuring this factor:
The Research Octane Number (RON) or the Motor Octane Number (MON). Irrespective of
which method is used, the higher the octane value the better the quality of the gasoline.
Kerosene – the principal use of kerosene is to produce jet fuel, which, similar to gasoline, will
have different formulations. In some parts of the world kerosene is used for cooking and
lighting.
Gas oils/Diesel distillate – these are used to produce diesel engine fuels and for households that
use oil as a source of heating.
Fuel oils – these are often used as a source of fuel for the power requirements of refineries and
power stations. They can also be used as fuel for ships, in which case it is termed ‘bunker
fuel’. (A bunker is simply a container in which the fuel for the ship is stored.)
Specialty products/Residuals – these include solids such as coke, asphalt tar, and waxes.
Typically, the lowest value products produced from the refining process.
166
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
volume for the same weight. Hence it is possible for production of refined products on
a per barrel basis to exceed that of crude oil. However, if the values are expressed on a
per tonne basis, one tonne of crude could never produce more than a tonne of refined
products.
For 2017, the US Energy Information Administrations (EIA) provided a breakdown
of the different products made from a single barrel:
▪ Gasoline – 20 gallons
▪ Ultra-low sulfur distillate (e.g. diesel fuels) – 11 gallons
▪ Other products – 6 gallons
▪ Jet fuel – 4 gallons
▪ Hydrocarbon gas liquids – 2 gallons
▪ Heavy fuel oil – 1 gallon
▪ Other distillates (e.g. heating oil) – < 1 gallon
On a national scale, the EIA produces annual statistics that show the type of products produced within the USA (Figure 6.3).
6.4.4
How does a refinery work?
No two refineries are the same and their complexity will vary according to their ability
to create a high-value product. The first stage of the refining process is fractional distillation. Crude oil is pumped through a furnace to release different liquids and gases,
Petroleum coke Refinery gases
8%
5%
Fuel oil
31%
Gasoline
46%
Jet fuel
10%
FIGURE 6.3 US refinery yields.
Source: EIA
167
Crude Oil
which are then fed into a distillation tower. Inside the tower the liquids and gases separate into different components (referred to collectively as ‘fractions’ or ‘cuts’) each with
a different boiling point. The component with the lowest boiling point evaporates first
and rises highest in the tower, whilst that with the highest boiling point evaporates last
and is collected at the base of the tower. The longer the carbon chain, the higher the
boiling point.
From the top of the distillation tower downwards the following components
(Table 6.2) are collected:
TABLE 6.2 Refined products from crude oil.
Temperature (∘ C)
Refined Product
Less than −45
−45–35
35–150
150–230
230–340
340–400
More than 400
Gas
Liquid Petroleum Gas
Gasolines and Naphtha
Kerosene
Gas oil
Heavy gas oil
Residue / Fuel oil
Once the crude has been separated into its fractions there is a second stage where
they are converted into intermediate products. Some of the fractions such as liquefied
petroleum gas may not require much conversion. Depending on the configuration of the
refinery, different fractions can be converted into more valuable products. For example,
if there is a high demand for gasoline, it can be produced from either naphtha or gas
oil, while diesel fuel could be made from either diesel distillate or gas oil. One of the
conversion techniques used is referred to as ‘cracking’ as it takes the heavy hydrocarbon
molecules (i.e crudes with high boiling points) and breaks them down into lighter products such as diesel and gasoline. Although there are several variants of the conversion
process (e.g. fluid catalytic cracking, hydro-cracking) the techniques use a combination
of heat, pressure, and a catalyst to achieve the required chemical reaction.
The third stage of the refining process purifies the product by removing pollutants
and contaminants. The final stage of the refining process involves modifying the intermediate products into the form in which they will actually be consumed. Different
fractions and other components are combined to produce finished products that have
specific properties. Products such as gasoline will have very tight formulation guidelines
that might be specified by law.
6.4.5
Refinery optimisation
Oil refiners function by converting crude oil into finished products with the aim of
maximising their net income. In the jargon of the refiner this is referred to as the
‘gross product worth’ or ‘gross production value’. However, to achieve this objective
168
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
the refinery must optimise their operations to get the biggest bang for their buck.
Optimisation will need to consider:
▪ Which types of crude oil the refinery can process – some refineries may be able to
process several crude oils or may be restricted to a very narrow range. This may be
a function of their location. For example, an inland refinery, served by a pipeline,
may only be able to take a single type of crude oil, whereas a coastal refinery may
have flexibility to receive a variety of different ship-transported oils.
▪ The yield from the crude – the proportion of each refined product that can be made
from a particular type of crude oil, which can be established by means of an assay.
▪ The configuration of the refinery – the way in which the refinery has been set up so
products can be made. For example, the refinery may choose different ‘cut points’
(the different temperatures at which the refined products are collected).
▪ Relative demand and supply for refined products – although refineries may lack complete flexibility to produce a diverse product slate, it may be possible to alter the
proportions of what is produced to meet market demand.
To illustrate the principles, consider the following simple example. Suppose a European refinery is configured such that it can accept two types of crude oil: Forties and
Oseberg. The oils are trading at USD 64.15 and USD 63.45 a barrel, respectively. The
refinery output (and their current market prices) consists of gasoline (USD 558.00/
metric tonne (MT)), jet fuel (USD 582.00/MT), gas oil (USD 615.00/MT) and fuel oil
(USD 350.00/MT). In order to optimise their production the refinery will need to know
what product yields can be derived from their source crude oils. A hypothetical set of
product yields is shown in Table 6.3.
As is common with refined products the unit of trading (e.g. metric tonnes) may be
different than that of the source crude (e.g. barrels). The following conversion factors
have been used to convert metric tonne prices to their barrel equivalent and based on
common market conventions1 .
Gasoline
Jet fuel
Gas oil
Fuel oil
8.35
7.88
7.46
6.35
TABLE 6.3 Hypothetical product yields from two types of crude oil.
Gasoline
Jet fuel
Gas oil
Fuel oil
1
Forties
Oseberg
35%
10%
20%
35%
25%
15%
15%
45%
The conversion factors are taken from BP’s annual statistical review of the energy markets which
can be accessed via their website.
169
Crude Oil
TABLE 6.4 Gross product worth and associated profit margins.
Gasoline
Jet fuel
Gas oil
Fuel oil
Gross product worth
Cost of crude oil
Profit margin
Forties
Oseberg
23.39
7.39
16.49
19.29
66.55
64.15
2.40
16.71
11.08
12.37
24.80
64.95
63.45
1.50
Table 6.4 Simplified calculation of a refineries gross product worth and overall profit
margin are expressed in USD. Operating costs are ignored for ease of illustration.
The gross product worth (GPW) and the associated profit margins are shown in
Table 6.4. For ease of illustration, all operating costs have been ignored.
The calculation for gasoline refined from a barrel of Forties is:
USD 558.00∕8.35 × 35% = USD 23.39
So, the income from gasoline on a per barrel basis is USD 66.82 (USD 558.00/8.35)
and 35% of a barrel of Forties is refined into gasoline resulting in revenue of USD 23.39.
The GPW is calculated as the sum of the revenue generated from the sale of the
refined products. The margin is the GPW minus the cost of the crude oil. Based on
this simple analysis, although Forties was trading at a higher price, the refiner should
logically choose this oil as it generates the greatest margin.
6.4.6
Refinery yields and relative crude oil prices
One of the features of the crude oil market is that most crude oils are traded as a differential to a benchmark oil such as Brent or West Texas Intermediate (WTI). These light
sweet crudes will tend to trade at a premium to heavy sour blends such as Urals (Russia)
or Maya (Mexico). In theory, the difference in price should be reflective of the different
value of the product slate produced from refining the different crudes. So if the refined
products produced from refining WTI return a value that is USD 5.00 higher than that
from the distillation of a grade such as West Texas Sour (WTS) Midland then it would
seem reasonable to assume the latter would trade at a USD 5.00 discount to the former.
Another way of looking at this differential is as a breakeven value. It represents the
discount that would need to exist such that the refiner could generate the same margin
from processing the benchmark crude as an alternative fuel. This would suggest that at
the discounted price the refinery would be indifferent between running the alternate
crude or the benchmark crude.
6.4.7
Measuring profitability
Generally, refining capacity has increased over the last few years (Figure 6.4).
170
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
120,000
Africa
S & C America
CIS
100,000
80,000
Middle East
Europe
60,000
N America
40,000
20,000
Asia Pacific
0
19
20
17
20
15
20
13
20
11
20
09
20
07
20
05
20
03
20
01
20
99
19
97
19
95
19
93
19
91
19
89
19
87
19
85
19
83
19
81
19
79
19
77
19
75
19
73
19
71
19
69
19
67
19
65
19
FIGURE 6.4 Global refining capacity 1965–2019. Thousands of barrels daily.
BP Statistical Review of World Energy 2020. BP PLC.
The figure shows that although the refining capacity of some countries has not
changed much (e.g. USA), other regions such as Asia Pacific have invested significantly.
It is very common for the reference to be made ‘refining margins’, which are calculated in a similar fashion to those illustrated in Section 6.4.5. Since it has been argued
that every refinery is different, it would be incorrect to say that there was a single margin value that would apply to the entire industry. So, it is common for price reporting
agencies to publish ‘indicator’ margins, which are essentially a generalised representation of the health of the downstream oil sector. They are more commonly referred to as
‘crack spreads’. In order to calculate these generic values, some assumptions have to be
made about how a ‘typical’ refinery in the respective region would be configured.
Figure 6.5 shows three such generic margins:
▪ US Gulf Coast sour cracking,
▪ Singapore medium sour hydrocracking,
▪ North West Europe light sweet cracking.
A 3-2-1 crack spread assumes three barrels of crude oil could be refined into
two barrels of gasoline and one barrel of fuel oil. This could be also be expressed as
1:0.67:33. Suppose that crude oil is trading at USD 50.00/bbl., while gasoline is trading
at USD1.40/gallon and heating oil at USD 1.20/gallon. Assuming there are 42 gallons
in a barrel, the crack spread is therefore the sum of the refined products minus the cost
of the crude oil.
= Gasoline + heating oil − crude oil
= (USD 1.40 × 42 × 0.67) + (USD 1.20 × 42 × 0.33) − USD 50.00
= USD 39.40 + USD 16.63 − USD 50.00
= USD 6.03
171
Crude Oil
$30
$25
$20
$15
$10
$5
$0
3Q
0
2Q 0
0
1Q 1
0
4Q 2
0
3Q 2
0
2Q 3
0
1Q 4
0
4Q 5
0
3Q 5
0
2Q 6
0
1Q 7
0
4Q 8
0
3Q 8
0
2Q 9
1
1Q 0
1
4Q 1
1
3Q 1
1
2Q 2
1
1Q 3
1
4Q 4
1
3Q 4
1
2Q 5
1
1Q 6
1
4Q 7
1
3Q 7
1
2Q 8
19
-$5
USGC Medium Sour Cracking
Singapore Medium Sour Hydrocracking
NWE Light Sweet Cracking
FIGURE 6.5 Refining margins 2000–2019. USD/bbl.
Source: BP Statistical Review of World Energy 2020. BP PLC.
It is also possible to measure product cracks. They would represent the added value
of converting, say, a barrel of heating oil into something more valuable such as gasoline.
6.4.8
Drivers of refinery performance and profitability
Almost by way of summary, the following factors would have an impact on refinery
profitability and performance. In very simple terms it is useful to consider those internal
factors over which the refinery has some control and external factors that are more
reflective of macroeconomic trends.
Internal factors
▪ The value of the product slate that is produced.
▪ The choice of crude oil to be used in the refining process.
▪ How the refinery is configured in relation to what type of crude oil it can process,
and which refined products are made.
▪ The reliability and efficiency of the plant.
▪ The location of the refinery will affect how much it costs to ship the freight and the
resultant products.
▪ The ability to produce specialty products with higher margins.
▪ Operating costs such as labour.
▪ Level of the refiner’s own inventories.
External factors
▪ Demand for refined products.
▪ The general level of inventories held by the market. Typically, there is an inverse
relationship between the level of inventories and price.
172
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Seasonality of demand.
▪ Crude oil prices.
▪ Ability to import or export products to manage a surplus or deficit.
▪ Possibility to substitute one fuel for another.
▪ Change in domestic product specifications.
▪ Advent of new technology.
▪ Change in relative demand (e.g. gasoline vs. diesel).
▪ Weather.
6.5
THE DEMAND FOR AND SUPPLY OF CRUDE OIL
‘ . . . as we know, there are known knowns; there are things we know we know.
We also know there are known unknowns; that is to say we know there are
some things we do not know. But there are also unknown unknowns – the
ones we don’t know we don’t know’.
—Donald Rumsfeld
6.5.1
Proved oil reserves
According to BP’s statistical review of world energy (2020), proved oil reserves are
‘generally taken to be those quantities that geological and engineering information
indicates with reasonable certainty can be recovered in the future from known
reservoirs under existing economic and operating conditions’.
Figure 6.6 shows the different regional proportions of proved reserves as of 2019,
while Figure 6.7 shows how these reserves have evolved over time.
Figure 6.6 shows that the vast majority of proved reserves exist in the Middle East.
However, the charts indicate that despite anecdotal concerns over the scarcity of crude
oil, the apparent level of reserves is increasing. At the end of 1980, the total recorded
figure was 684 thousand million barrels (tmb) but had risen to 1,734 tmb by the end of
2019.
It may seem puzzling to hear stories of declining oil reserves and then see such
contradictory data as presented here. However, the definition of reserves would suggest
that an increase in price and/or the implementation of new technology might account
for the increase.
6.5.2
R/P Ratio
The Reserves to Production (R/P) ratio indicates the length of time (in years) that a
country’s remaining reserves would last if production were to continue at current levels.
It is calculated by dividing the reserves remaining at the end of a year by the production
in that year. The data used in the first edition of the book (2005) indicated that on a
global basis, the values would suggested there would be sufficient reserves for another
41 years of consumption at production levels that prevailed at the time. Using the 2019
figures, that value has increased to about 50 years. This means that the number has
173
Crude Oil
Europe
4%
Africa
13%
Middle East
23%
CIS
8%
N America
8%
S & C America
44%
FIGURE 6.6 Relative proved reserves by region 2019.
BP Statistical Review of World Energy 2020. BP PLC.
2000
Europe
Asia Pacific
Africa
1800
1600
CIS
1400
N America
1200
1000
S & C America
800
600
400
Middle East
200
0
18
20
16
20
14
20
12
20
10
20
08
20
06
20
04
20
02
20
00
20
98
19
96
19
94
19
92
19
90
19
88
19
86
19
84
19
82
19
80
19
FIGURE 6.7 Evolution of proved reserves 1980–2019. Thousand million barrels.
BP Statistical Review of World Energy 2020. BP PLC.
174
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Asia Pacific
Africa
Middle East
CIS
Europe
S & C America
North America
0
20
40
60
80
2004
100
120
140
160
2019
FIGURE 6.8 R/P ratios in 2004 compared to 2019.
Source: BP Statistical Review of World Energy 2020. BP PLC.
to be treated with some caution and is only a snapshot at any point in time. It may
change if new reserves are found or if demand for oil declines due to an increased use
in alternative sources of energy.
Figure 6.8 shows how the R/P ratio has changed on a regional basis when comparing 2004 with 2019.
6.5.3
Production of crude oil
Figure 6.9 shows that between 1965 and 2019, production of crude oil rose from an
annual figure of 31,799 in 1965 to 95,192 thousand barrels daily (tbd) by 2019.
Figure 6.10 illustrates the different proportions of production on a regional basis
as of the end of 2019. One of the simple messages to highlight is that there is a big
difference between the location of the reserves and production. So, although South and
Central America have 19% of the world’s proven reserves, they only account for 6% of
global production.
In Section 6.4.5 it was argued that most crude oils are priced relative to a particular
benchmark. Peak production for one of the key benchmarks (Brent, North Sea oil) was
reached in 1999 with a total of 2,909 thousand barrels per day (mbd), which made up
about 2.5% of the total world production. By the end of 2019, this figure had fallen to just
1,118 mbd, amounting to just over 1% of world production. This has called into question
whether this particular stream of crude oil is truly representative of the global demand
and supply trends.
Figure 6.10 also shows the impact of shale oil on the US market as growth rates for
US crude production in the decade 2008–2018 were 8.5%; this after a period in which
US production rates had been falling.
175
Crude Oil
120,000
100,000
Europe
S & C America
Asia Pacific
80,000
Africa
60,000
CIS
40,000
N America
20,000
Middle East
0
19
20
17
20
15
20
13
20
11
20
09
20
07
20
05
20
03
20
01
20
99
19
97
19
95
19
93
19
91
19
89
19
87
19
85
19
83
19
81
19
79
19
77
19
75
19
73
19
71
19
69
19
67
19
65
19
FIGURE 6.9 Evolution of crude oil production between 1965 and 2019. Thousand barrels daily.
Source: BP Statistical Review of World Energy 2020. BP PLC.
S & C America
6%
Europe
4%
Asia Pacific
8%
Middle East
32%
Africa
9%
CIS
15%
N America
26%
FIGURE 6.10 Relative crude oil production by
region.
Source: BP Statistical Review of World Energy 2020.
BP PLC.
176
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
6.5.4
Consumption of crude oil
Figure 6.11 shows the consumption of crude oil over time on a regional basis, while
Figure 6.12 presents the proportion of regional consumption for the end of 2019.
The significant facts are:
▪ The consumption of crude oil has continued to rise globally.
▪ Asia-Pacific is the largest single consuming region, accounting for 37% of the total
world consumption.
▪ China continues to consume more oil on an absolute and percentage basis. For
example, in 2005, China consumed about 8.5% of world production. By 2019, it
consumed about 14% of a greater overall consumption amount.
▪ There is a mismatch between the location of reserves, the amount produced, and
the amount consumed. So, although the Middle East has 23% of proved global
reserves, it produces over 32% of the world’s crude oil and only consumes 10% of
the total.
6.5.5
Crude oil trade
One of the themes highlighted in the previous section is that the location of oil reserves,
where it is produced and where it is consumed can differ significantly. This gives an
120,000
100,000
80,000
Africa
CIS
S & C America
Middle East
Europe
60,000
N America
40,000
20,000
Asia Pacific
0
19
20
17
20
15
20
13
20
11
20
09
20
07
20
05
20
03
20
01
20
99
19
97
19
95
19
93
19
91
19
89
19
87
19
85
19
83
19
81
19
79
19
77
19
75
19
73
19
71
19
69
19
67
19
65
19
FIGURE 6.11 Evolution of crude oil consumption on a regional basis. 1965–2019. Thousand
barrels daily.
Source: BP Statistical Review of World Energy 2020. BP PLC.
177
Crude Oil
S & C America
6%
CIS
4%
Africa
4%
Asia Pacific
37%
Middle East
10%
Europe
15%
N America
24%
FIGURE 6.12 Relative consumption of crude oil 2019.
Source: BP Statistical Review of World Energy 2020. BP PLC.
insight into the nature of the international trade for crude oil. Figure 6.13 shows some
of the regional flows of crude oil. For presentation purposes the diagram shows the
main regional flows of crude oil and illustrates how certain areas (e.g. West Africa and
the Middle East) will transport crude oil to those regions with the greater demand (USA,
Europe, and China).
6.5.6
Demand for refined products
The BP statistical review provides a regional breakdown of the demand for the refined
products (Figure 6.14). Three of the categories are:
▪ Light distillates – aviation and motor gasolines and light distillate feedstock.
▪ Middle distillates – jet and heating kerosenes; gas and diesel oils.
▪ Fuel oil – marine bunkers and crude oil used directly as fuel.
Perhaps somewhat unsurprisingly, the demand for light and middle distillates dominates all the other categories.
178
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
FIGURE 6.13 Inter-regional movements of crude oil. Million tonnes.
Source: BP Statistical Review of World Energy 2020. BP PLC.
Africa
CIS
Middle East
S & C America
Europe
N America
Asia Pacific
0
2,000
4,000
Fuel oil
6,000
8,000
Middle distillates
10,000
12,000
14,000
Light distillates
FIGURE 6.14 Regional consumption of different refined products. Thousand barrels daily.
Source: BP Statistical Review of World Energy 2020. BP PLC.
179
Crude Oil
6.5.7
Security of supply (and demand)
One of the key themes in most energy markets is the security of supply. This can be
interpreted in a number of ways but at a simple level if can be thought of as the ability
of any country to ensure there is a sufficient ongoing supply to meet expected demand.
One of the ways in which this can be achieved is by diversifying the sources of supply.
The largest individual crude oil importers as of 2019 are:
▪ Europe: 522 million tonnes
▪ China: 507 million tonnes
▪ USA: 338 million tonnes
For each of those countries the three main sources of supply are (in descending
order of volume):
▪ Europe: Russia, Other CIS, West Africa;
▪ China: Middle East, West Africa, Russia;
▪ USA: Canada, South and Central America, Saudi Arabia.
Likewise, producers of crude oil are equally sensitive to changes in demand patterns and should therefore have several different supply destinations. Examples of a
concentration of supply include Russia, whose main export location is Europe.
6.6
PRICE DRIVERS
‘Oil is not ours, it’s theirs. You don’t find oil in Switzerland; when prices go up,
everyone wants to share the bonanza’.
—Paolo Scaroni, Chief Executive of ENI
There are several factors that drive the price of crude oil. For convenience, the issues
have been collated under three generic headings: supply chain considerations, geopolitics (i.e. the study of relationships between nations), and macroeconomic issues. In
addition, there will be a series of factors that influence the demand and supply for the
various refined products, which may feed back to the price of crude oil. Ultimately, the
market for crude oil is demand led with money flowing to the producing countries.
6.6.1
Macroeconomic issues
Economic Activity
If there is strong growth in Gross Domestic Product (GDP), signifying increased economic activity, then it is likely that the price of crude oil will increase accordingly.
This may also have another effect in that if consumers believe that prices will be
higher in the future, they may be tempted to buy forward. Sellers on the other hand will
be happy to enjoy the benefits of a higher price and may remain unhedged.
180
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Reserves and production
Arguably, the most contentious question that arises with respect to crude oil is how
much is left? Undeniably, crude oil is a finite resource, which at some point will be
exhausted.
When discussing this issue, reference is often made to the concept of Hubbert’s
peak. Hubbert, a US geologist, predicted in the 1950s that US crude oil production would
peak in the early 1970s; according to BP’s annual energy review, an apparent peak was
reached in 1972 (11.1 m barrels a day). Thereafter US production declined until 2008
(6.8 m barrels a day). However, the introduction of shale oil production led to a substantial increase and by the end of 2019, the US was producing just over 13 m barrels daily.
On a global scale after a decline in the early 1980s production has generally increased
(see Figure 6.15).
For existing wells, the issue is more of how to extract what remains, and as a rule
of thumb, reservoir recovery rates are about 30%, but this could vary considerably
(e.g. 10–70%). At a very simple level it has to be economically viable for a producer
to recover what is remaining. Although this will be a function of several variables, it
would seem reasonable to suggest that a producer has to earn an acceptable return on
capital invested to make such extraction worthwhile.
This means that the effectiveness of technology to retrieve the oil and the price they
will receive for it plays an important role. There will also be known reserves that are
difficult to access but with improved technology, and an increasing price, may become
worthwhile to extract. Then there is ‘unknown’ technology where economic circumstances may act as a spur for further innovation that is at present inconceivable. It also
may be that there are large reserves that are as yet undiscovered. Despite the oft made
remark that the world is getting smaller, there are large areas of many existing oil producing countries that remain un-surveyed in terms of potential oil production. As a
result, companies will continue to search for oil in the hope that they discover a major
reservoir, sometimes referred to in the industry as an ‘elephant field’.
120,000
100,000
80,000
60,000
40,000
20,000
0
19
20
17
20 5
1
20 3
1
20
11
20 9
0
20 7
0
20
05
20 3
0
20 1
0
20 9
9
19
97
19 5
9
19 3
9
19
91
19 9
8
19 7
8
19
85
19 3
8
19 1
8
19 9
7
19
77
19 5
7
19
73
19 1
7
19
69
19 7
6
19 5
6
19
FIGURE 6.15 Global oil production. 1965–2019. Thousand barrels daily.
Source: BP Statistical Review of World Energy 2020. BP PLC.
Crude Oil
181
Availability of Strategic Reserves
The USA maintains a Strategic Petroleum Reserve, which was designed to provide the
country with an emergency supply of crude oil. It is a US government complex that
consists of four sites with deep underground storage caverns created in salt domes along
the Texas and Louisiana gulf coast. The reserve can hold some 727 million barrels.
Possibility of substitution
One of the effects of the substantial rise in oil prices in the 1970s was that it ultimately
encouraged oil users to switch to alternative sources, notably to natural gas. This movement away from crude oil was partly responsible for the low crude oil prices of the
1980s. With the prospect of many carbon fuels being eventually exhausted and 65% of
each barrel of oil being consumed by the transport sector, many oil companies are now
attempting to diversify into alternative sources of energy. As the price of crude oil rises
and concerns over climate change increase, there is an incentive to switch to alternatives such as biofuels (ethanol), gas-to-liquids technology (a process that can be used
to turn gas or coal into products normally produced from crude oil), hydrogen fuel and
fuel cells, solar energy, nuclear power, and wind and wave technology.
The use of hydrogen as a replacement for gasoline has been mooted as a substitute
due to the lack of side effects. Hydrogen is added to a fuel cell and mixed with oxygen,
which chemically reacts to produce electricity. This powers an electric motor, which
propels the car. The resultant exhaust comprises nothing more than water or vapour.
Even this technology is not without problems as one of the likeliest sources of hydrogen
is expected to be natural gas. To date production of the cars in significant numbers has
not transpired to date. A major obstacle to its development is the significant cost of the
attendant global refueling infrastructure.
A rising oil price may encourage users to seek out alternative products, but it may
also encourage companies to look for oil in remote parts of the world and even in
deep-water areas. One of the more unconventional areas of development is the Canadian oil sands of Alberta. Oil sands generally refer to mixtures that have the consistency
of molasses but may comprise several organic materials such as bitumen. Bitumen has
naphthenic properties in that it has a high viscosity and is not particularly flammable.
The API gravity of the bitumen found in the Canadian oil sands varies between about
7–13, while the sulfur content is 4–6%. Once recovered from the earth, it is upgraded to
heavy crude oil before being transported along pipelines to a refinery where it can be
processed and refined. The attractiveness of the project is that it offers a source of crude
oil supply many times greater than that of Saudi Arabia.
However attractive this may seem, there are some drawbacks. Two tonnes of oil
sands yield about 1.25 barrels of bitumen and a single barrel of crude. Secondly, the
energy required for extraction is considerable and would consume a large proportion
of the country’s natural gas output. The extraction process also requires a significant
amount of water and there are the related issues of origin, storage, and transport. In
addition, the environmental issues of restoring the countryside to its natural state can
no longer be ignored.
182
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Price of other crude oils
Since most crude oils are priced as a differential from a series of core marker crudes
such as WTI and Brent, movements in one of the market crudes will have a knock-on
effect to the remaining market
Demand for refined products
Each refined product will have its own unique supply and demand dynamics and in
some cases, these may feed back into the crude oil market. For example, in winter crude
oils that are favoured to produce heating oils will gain in value, while a similar effect is
seen in the summer for the crude oils that yield larger amounts of gasoline.
Taxation
In the late twentieth century, European governments gave drivers substantial tax incentives to buy diesel-powered cars partly in the belief they emitted fewer greenhouse gases.
This led to an increase in the proportion of diesel cars sold in Europe and as a result,
consumption of diesel increased accordingly. However, the increase in demand was not
matched by a similar investment in refining technology that would allow the conversion
of heavier fuel oils to the more attractive middle distillates such as diesel. As a result,
diesel prices rose above those for gasoline. At the time of writing, there has been some
negative publicity over the environmental impacts of diesel, which has subsequently
led to a reduction in demand for this type of vehicle. Similarly, some governments have
introduced subsidies to encourage drivers to switch to electric vehicles.
The pump price of gasoline is made up of several components:
▪ The cost of the crude oil.
▪ The refinery’s margin.
▪ Taxation.
▪ Sales margins.
The taxation of gasoline is always a politically sensitive subject, which in some
countries has given rise to civil unrest. The level of taxation could have an impact on
consumer behaviour, which could feasibly feed back along the supply chain.
Transportation trends
This could manifest itself in several ways:
▪ A shift away from diesel cars.
▪ An increase in demand for electric vehicles (EVs).
▪ Changes in regulation that impact the technical specification of a particular type of
fuel (e.g. sulfur limits in fuel oil).
For example, if there is an increase in EVs, then one might expect a decrease in the
demand for gasoline and hence crude oil and an increase in demand for those fuels that
are required to produce electricity (e.g. coal and natural gas).
Crude Oil
183
Investor and speculative activity
A key theme over the last few years has been changing interest shown in commodity
markets by financial participants. Given the size of their interest and money flowing
into the market, their activities are often cited as being responsible for creating both
upward and downward price spirals.
Since most investors would be reluctant to take delivery of the physical commodity, the most popular method to take exposure to commodities is by the futures market.
However, some funds may be constrained in their ability to transact futures and so
exposure is taken via a total return transaction. Here the investor enters into an agreement with an investment bank where they receive a cash flow that mimics the return
on a commodity index such as the S&P GSCI®. This index takes its value from the
prices of several commodity futures and has a significant crude oil weighting. Although
the nature of the index will be considered in more detail in a subsequent chapter, the
‘rolling’ of futures contracts generates part of the return to an investor. The futures roll
describes the process of selling the front month contract (the ‘prompt’ contract) as it
approaches maturity and buying the next delivery month to maintain exposure to the
commodity. With significant investor inflows into the index, this activity can be significant and may result in an element of distortion in the shape of the short end of the
forward curve.
Investment in the S&P GSCI is essentially passive in nature as the index is compiled
according to a predetermined set of rules. It is attractive to ‘real money’ accounts such
as pension funds and mutual funds as it represents an unleveraged ‘long only’ futures
investment. Active investors (such as hedge funds or commodity trading advisors) may
consider an index approach somewhat restrictive and seek to earn an enhanced return
that exceeds the performance of the index. Index investors will earn a return equal
to the market, which is sometimes referred to as ‘beta’. Active investors try to outperform the index, the enhanced return being referred to as ‘alpha’.
Speculative activity has long been seen as some form of evil that causes commodity prices to move from some notion of ‘fair value’, which of course is never actually
revealed. This takes us back to the issues of price formation, which were considered in
Chapter 2 using the toddler and robot argument. The author feels that it is fair to say
that it is likely that speculative flows may have some impact on prices, but it would be
incorrect to say that they are always the dominant factor. If they were, then all the other
factors influencing price would have to have very little impact, which is unlikely.
Movement of the USD
Like most commodities, crude oil is priced in US dollars, and so a weakening of the
currency should lead to an increase in the demand for the commodity from non-dollar
users, as the cost is now lower in domestic currency terms.
Weather
Weather and other seasonal factors will also play an important role in the balance
of demand and supply. An obvious example would be the increase in demand for
184
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
heating oil during winter. The US ‘driving season’, which extends roughly from June–
September is a period where demand for gasoline increases as people take vacations.
Another impact of weather is that very severe events such as hurricanes may impact
production and refining capabilities.
6.6.2
Supply chain considerations
Upstream production
This is the capacity at the point of (on- or offshore) extraction. Some large producers
have chosen to keep a certain amount of idle capacity to respond to any sharp increase in
demand. Typically, Saudi Arabia has retained sufficient capacity to meet any temporary
upswing in demand, but this should not suggest that they have an infinite capacity to be
able to perform this function. Indeed, their spare production capacity is in the heavier
crudes with higher sulfur content. This will only be of use if there is sufficient refining
capacity for this type of crude oil.
Like many industries, the impact of technology will have an impact on production.
Generally, only about 30% of oil is recovered2 from a particular well as it is not always
cost effective. As technology improves, it is likely that the amount of oil that can be
extracted will increase.
Costs of production
There are three main categories of cost in producing a barrel of oil. Exploration costs
capture the cost of finding the oil (e.g. exploration and appraisal drilling). Development
costs include the procurement of equipment, facility construction, drilling, engineering, and project management. Operating costs are essentially the day-to-day expenses
of running the site. The Wall Street Journal (2016) looked at the average cash cost to
produce a barrel of oil, which varied from about USD 9.00/bbl. for Saudi Arabia to USD
44.00/bbl. for the UK.
Oil field Activity
Market participants closely monitor current production to identify any possible supply disruptions. For example, a monthly count of all rotary rigs is published, while the
maintenance of existing fields is also closely scrutinised.
Refining capacity
If there is a sudden increase in demand, there may not be sufficient physical refining
capacity resulting in an increase in the price of oil. The construction of oil refineries has
a very long lead-time, is expensive, and must overcome many environmental considerations. For example, expansion of an existing facility could take at least two years while
2
Recovery rates can vary considerably from about 10–80%.
Crude Oil
185
construction of a new refinery, upwards of five years. In the USA in 1981, there were
325 refineries with a total capacity of 18.6 million barrels a day. By late 2005, there were
148 refineries, and by 2017 this fell to 135. Global estimates suggest there are about 700
facilities.
Environmental considerations
As many governments increase the environmental regulations the possibility of building large infrastructure projects decreases. Take the construction of a refinery in the
USA; from the mid-1990s onwards refiners spent some USD 47 billion meeting the
demands of a plethora of environmental laws. According to the Energy Information
Administration (EIA) there were no refineries built between 1998 and 2014 in the USA.
Refinery yields and margins
One of the key price relationships within the crude oil market, which measures the relative attractiveness of the refined products to crude oil, is the crack spread. This measures
the difference between the income generated by the refined products and cost of the
crude oil used in the refining process. This aspect of refinery profitability is considered
in Section 6.4.7.
If the oil market experiences a general fall in supply, those refineries that can only
process lighter crudes will bid up the price of this type of oil. This may result in the price
differential between light and heavy crude oils moving beyond its theoretical value3 . In
a rising price environment a complex refinery that has the ability to take the heavier
and more sour crude oils and refine them into higher value products will be able to
generate greater margins as these types of crude will tend to trade at a higher discount
to the lighter, sweeter crudes.
Storage capacity and inventories
When asked what key variable he looked at every day, the veteran oil trader Gary Ross
declared it all boiled down to one thing. ‘What is the demand for inventory? It is in the
mind of everyone who buys and sells oil . . . and that’s driven by psychology’. (Financial
Times, 2019)
According to the EIA, US storage capacity in 2018 was just over 710 million barrels, and given it is finite its availability will have an impact on price. Take a situation
where there are concerns about the future security of crude oil supplies, encouraging
participants to build their crude oil inventories. Storage should become more expensive,
driving up the cost of oil for forward delivery. However, if there is nowhere to store the
crude oil, producers may eventually be forced to cut production, increasing the cost of
shorter-dated crude oil relative to longer-dated maturities.
Another related statistic relates to the relative utilisation of the capacity, which captures the current level of inventories. A high level of storage utilisation would indicate
a high level of inventories relative to demand, which may have a negative impact on
price.
3
See 6.4.6 for an explanation of crude oil differentials.
186
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Availability of supporting resources
Although much is made of the production/refining/consumption triangle, an
often-overlooked feature of the supply chain is the availability and cost of such items
as oil rigs, tankers, and a skilled workforce. A rise in commodity prices would increase
the cost of raw materials such as steel, thus raising the cost of large projects such as oil
rigs and refineries. This has a knock-on effect in that it increases the cost of producing
a barrel of oil.
One interesting example relating to the cost of associated services is the cost of shipping oil. In mid-2006, BP announced the shutdown of crude oil production at Prudhoe
Bay in Alaska due to maintenance issues. At the time, the field was producing about
400,000 barrels of crude oil every day and was a major source of supply for refineries
on the West Coast of the USA. To meet its existing commitments BP was forced to look
overseas at alternative sources of supply. Estimates4 suggested that it would take 12 very
large crude carriers (VLCCs) over a 60-day period to transport the deficit in production
to the USA. In a market where the number of vessels capable of moving this amount
of crude oil was estimated at 400 to 450, the extra demand for the freight would move
charter prices significantly and have a knock-on effect on margins.
Infrastructure spending
In theory, a rising oil price should encourage producers to spend more money on
improving the existing infrastructure and to invest in new production opportunities.
However, producers have very bad memories of crude oil priced at USD 10.00/bbl. and
are often reluctant to invest in times of high oil prices in case of a subsequent fall.
Distribution
This relates to the movement of crude oil along the supply chain. Markets such as the US
are generally pipeline based, but Europe will tend to be more waterborne. So relevant
issues include:
▪ The demand for and supply of tankers,
▪ Seaborne disruptions such as piracy,
▪ Port capacity and berth availability,
▪ Pipeline capacity and associated demand.
Natural Disasters
In a similar vein, natural disasters that lead to shutdowns of rigs, entire fields, or refining
capacity will have a substantial impact on the price of crude oil. For example, Hurricane Katrina hit the US Gulf Coast in the summer of 2005. This led to the closure of
approximately 10% of the nation’s refining capacity and 90% of the US Gulf of Mexico output. Shortly afterwards, Hurricane Rita led to a temporary shutdown of 27% of
refining capacity.
4
Financial Times, (15th August 2006) The dramatic knock on effect of BP’s Prudhoe shutdown
187
Crude Oil
Quality of Crude Oil
The quality of oil from a particular field is not necessarily constant. Since oil extracted
from different wells within a designated area may be collected into a single system, the
quality of each constituent crude oil may be different.
6.6.3
Geopolitics
War, Terrorism, and Sanctions
Although the motives for the major Middle East conflicts are outside the scope of this
book, the impact on production and consequently price of crude has been significant.
A country with substantial reserves such as Iraq will influence global production when
physical capacity to produce and deliver is destroyed. Figure 6.16 shows that after the
First Gulf War in 1991, Iraq production fell significantly and took many years to recover.
Associated with the various conflicts have been acts of terrorism in the Middle East and
East African region. One concern relating to a potential terrorist attack is Saudi Arabia
where two-thirds of production is moved through two processing plants and a single
terminal. The imposition of sanctions on countries such as Iran would also have price
implications.
Internal unrest
Oil rich countries such as Venezuela and Nigeria have suffered bouts of internal unrest
that have threatened to reduce crude oil production in their respective countries.
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
19
20
17
20
15
20
13
20
11
20
09
20
07
20
05
20
03
20
01
20
99
19
97
19
95
19
93
19
91
19
89
19
87
19
85
19
83
19
81
19
79
19
77
19
75
19
73
19
71
19
69
19
67
19
65
19
Iran
Iraq
FIGURE 6.16 Iran and Iraq production 1965–2019. Thousand barrels daily.
Source: BP Statistical Review of World Energy 2020. BP PLC.
188
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Political tensions and the security of supply
Reference is often made to the ‘fear factor’ within commodity markets and usually in
relation to concerns over the security of supply. There are ongoing concerns relating to
the nuclear capability of Iran. As China continues to industrialise, it has shown that it is
prepared to secure supplies from countries with which the USA has refused to deal (e.g.
Iran, Sudan, Angola). This has led some commentators to speculate on future tensions
between the superpowers as they both wrestle for control of strategic oil supplies.
It is sometimes assumed that the suppliers of crude oil exert more influence over the
price than the buyers. However, this is not always the case, and it is valid to consider
the notion of ‘security of demand’. That is, oil-producing companies must be certain
that they will be able to find a buyer for their output. This means that while buyers are
looking for diversification of supply, producers are looking for certainty of demand.
Resource Nationalism
In some oil countries the substantial price rises of the early twentieth century led to
an increase in nationalism as newly elected governments threatened to nationalise oil
production in their respective countries (e.g. Bolivia, Ecuador, Venezuela). Russia has
also reasserted control over some aspects of energy production as prices have risen.
However, resource nationalism may manifest itself in less obvious forms as for example,
at the time when concerns were being raised about nationalisation in South America,
the UK increased taxes on revenues earned by oil companies operating in the North Sea.
Access to new reserves
National (i.e. government owned) oil companies own a significant portion (about 75%)
of the world’s oil. These entities may choose to allow international oil companies access
to the oil where they themselves do not possess a certain technical capability. This has
made it difficult for private oil companies to replenish their reserves and has driven
them into new areas of research (e.g. gas-to-liquids technology) or risky areas of exploration (e.g. Canadian oil sands or 10,000-foot-deep waters off Mexico).
Organisation for Petroleum Exporting Countries (OPEC)
According to the organisation’s website their mission statement is:
‘ . . . to coordinate and unify the petroleum policies of Member Countries and
ensure the stabilisation of oil prices in order to secure an efficient, economic
and regular supply of petroleum to consumers, a steady income to producers
and a fair return on capital to those investing in the petroleum industry’.
The current list of the 14 OPEC members:
▪ Algeria
▪ Angola
▪ Congo
▪ Ecuador
189
Crude Oil
▪ Equatorial Guinea
▪ Gabon
▪ Islamic Republic of Iran
▪ Iraq
▪ Kuwait
▪ Libya
▪ Nigeria
▪ Saudi Arabia
▪ United Arab Emirates
▪ Venezuela
OPEC meets twice a year and the main technique used to achieve their objectives
is to target output quotas for each of the member countries. Although the countries
collectively do not produce the entire global output of oil (they are believed to control
about 37%5 ), their output is significant, and the market closely monitors their activities. Based on the different price factors presented in this section, it would be perhaps
unrealistic to say that OPEC controls the price of oil, but its prominence in terms of
overall production cannot be ignored. The relative power that they can exert would be
greater if the market was relatively short of oil, but with the advent of new sources such
as shale, this influence is arguably lower. This has given rise to the idea of the ‘OPEC
Call’ – if non-OPEC members require additional supply, then it could in theory be met
from OPEC production.
6.6.4
Analysing the forward curve
So far, the analysis has concentrated on generic price factors without necessarily considering their term structure. Davis argues that the term structure of oil can be broken into
two segments: 0–18 months and beyond 18 months. He argues that the 0–18-month
segment is closely linked to the physical market and reacts mostly to issues such as
demand and supply, the level of inventories, availability of storage, and security of supply. Beyond those maturities, market activities focus more on financial rather than physical concerns; this may include expectations over interest rates and inflation. It is also
where market participants are likely to express their views on proposed capital expenditure and anticipated infrastructure projects.
He also notes that the term structure is more volatile than financial markets, may
display seasonality, and that crude oil prices tend to display mean reversion over the
longer term.
6.7
THE PRICE OF CRUDE OIL
6.7.1
Defining price
The price of crude oil is expressed in US dollars per barrel, sometimes shortened to
USD/bbl. The origin of the abbreviation ‘bbl.’ is unclear with two competing explanations. One story argues it derived from the practice of storing crude oil in blue coloured
5
BP Annual Statistical Review of Energy 2020
190
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
barrels, whereas some argue that the abbreviation arose from the fact that beer barrels
were originally used for storage. A barrel will hold 42 US gallons or 159 litres.
6.7.2
The evolution of crude oil prices
High crude oil prices are often seen as a modern-day problem. However, if one adjusts
for the effect of inflation a very different picture emerges.
Figure 6.17 indicates that when adjusted for inflation there are two significant price
peaks. The first was in 1864 (an early ‘boom and bust’ period in the oil industry) when
the price exceeded USD 100.00/bbl. and in 1980 when the price nearly touched USD
110.00 (the first global oil crisis).
One of the interesting questions about high oil prices is whether they will trigger a
global recession. If the price rise has been caused by a restriction in the supply of crude
oil, there may be a tendency for inflation to rise and output to fall. But if there has been
an increase in demand brought about by a significant increase in productivity growth
in oil importing countries, or increased spending in the oil exporting countries then a
high oil price could be consistent with greater economic expansion.
6.7.3
Delivered price
What will be delivered and the nature of the price to be paid must be specified on the
contract, which typically will include the following:
▪ Price (usually expressed in USD/bbl.).
▪ Whether the contract is to be priced off an index or a‘marker’ crude.
▪ Any differential that should be applied to the price.
140
120
100
80
60
40
20
0
16
20 1
1
20 6
0
20 1
0
20
96
19 1
9
19 6
8
19
81
19 6
7
19
71
19 6
6
19
61
19 6
5
19
51
19 6
4
19
41
19 6
3
19 1
3
19
26
19 1
2
19 6
1
19 1
1
19 6
0
19
01
19 6
9
18 1
9
18
86
18 1
8
18
76
18 1
7
18 6
6
18 1
6
18
$ money of the day
$2019
FIGURE 6.17 Price of crude 1861–2017; USD/bbl. Expressed as inflation adjusted (USD 2019)
or in the money of the day.
Source: BP Statistical Review of World Energy 2020
191
Crude Oil
▪ Quality (e.g. sulfur content).
▪ When it will be delivered.
▪ Where it will be delivered.
▪ Whether the price is CIF or FOB.
The price of crude oil is usually expressed in one of two ways: Free on Board (FOB)
or Cost, Insurance, and Freight (CIF). An FOB price is the most common pricing
method as it represents the price of the commodity at the point of loading and therefore
allows comparisons to be made between different crude oil prices around the world.
However, a purchaser of crude is going to be more interested in the CIF price as it gives
an indication of the total cost.
Typically, the cost of delivering crude oil to a purchaser will include:
▪ The quoted FOB price.
▪ Shipping costs.
▪ Any associated secondary transportation expenses such as pipeline costs (if the
refinery is not based on the coast).
▪ Losses due to factors such as evaporation.
▪ Insurance costs.
▪ Cost of financing oil purchased while in transit and prior to product production
(there will be a gap between when oil is paid for and when the refined products are
sold to generate income).
6.7.4
Marker crudes
In Section 6.2 it was noted that the price reporting agencies often quote prices for over
100 different types of crude oil. Section 6.4 noted that many crude oils were quoted as a
differential to a so-called ‘marker’ crude. In theory this differential is a reflection of the
discount (or premium) that should be applied to a particular grade of crude oil relative
to another crude such that a refiner would be indifferent in using either to produce a
given product slate.
The key global marker crudes that are used as pricing benchmarks are:
▪ West Texas Intermediate (USA origin; global benchmark)
▪ Light, sweet crude oil with an API of 39.6 and a sulfur content of 0.24%
▪ Brent (North West Europe origin; global benchmark)
▪ Light, sweet crude with an API of 38.8 and a sulfur content of 0.37%
▪ Tapis/Oman (Malaysian/Middle East origin; Asia Pacific benchmark)
▪ API of 46 and a sulfur content of 0.03%
▪ Urals (Eastern Europe origin; Mediterranean benchmark).
▪ API of 31.7 and a sulfur content of 1.35%
The selection of a crude oil as a benchmark does not follow any predetermined
formula. For example, Brent is a key global benchmark, even though it now only constitutes about 0.3% of total global output. However, an advantage is that it originates from
a politically stable region and is supported by an actively traded derivatives market, like
192
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
West Texas Intermediate (WTI) in the USA. Even within a single type of crude oil there
may be different benchmarks. For example, the price for physical Brent crude (‘Dated
Brent’) may be used to price certain commercial contracts whereas other participants
may prefer Brent futures.
Other benchmarks may appear for different reasons. The use of Light Louisiana
Sweet has become a useful domestic US benchmark for Gulf Coast refiners, with Mars
being a popular domestic sour crude benchmark.
However, even the global markers such as WTI may be subject to unpredictable
behaviour. The CME crude oil future is a physically settled crude oil with the delivery
point in Cushing, Oklahoma. The supply and demand for crude in this area can have an
impact on its price. So, if there were, say, an unplanned outage in a local refinery there
would be knock-on effect that would increase the level of inventories and put downward
pressure on price. As a result, some consider WTI to be more of a ‘land locked’ crude oil
not representative of global conditions but perhaps more relevant to US mid-continent
producers and refiners.
The price of crude oil referred to by OPEC is based on a basket of oils (ORB – OPEC
Reference Basket) produced by the member countries. The ORB is made up of the following crudes:
▪ Saharan Blend (Algeria)
▪ Girassol (Angola)
▪ Djeno (Congo)
▪ Oriente (Ecuador)
▪ Zafiro (Equatorial Guinea)
▪ Rabi Light (Gabon)
▪ Iran Heavy (Islamic Republic of Iran)
▪ Basra Light (Iraq)
▪ Kuwait Export (Kuwait)
▪ Es Sider (Libya)
▪ Bonny Light (Nigeria)
▪ Arab Light (Saudi Arabia)
▪ Murban (UAE)
▪ Merey (Venezuela)
However, OPEC countries are free to use different approaches to price their own
physical contracts. For example, Saudi Arabia will price its oil depending on its destination. So, if the crude is going to the USA, it is priced as a differential to WTI or to the
American Sour Crude Index; crude going to Europe is priced as a differential to Brent.
It may seem surprising that given Saudi Arabia holds a significant proportion of the
proven reserves of crude oil, as well as being a key supplier, that its crude oils are not
considered as global markers. The reason for this dates to the 1980s when the crude oil
producers introduced ‘net back pricing’. Producing nations rationalised that the value of
crude oil lay in the different products that could be produced from the refining process.
At the time, the refined products were trading at relatively high prices and so the producers set the price of crude based on these higher values, less a margin for refining and
transportation. Under this system, the refiners would earn a fixed margin irrespective
193
Crude Oil
of how much they produced and so were incentivised to run at high capacity. This led to
an oversupply of products, whose prices fell significantly as well as those for crude oil.
As a result it is more conventional for Saudi crude oil to be priced according to the terminal location or according to the date of final delivery (e.g. 40 days after loading – the
time taken to transport the crude oil).
One of the consequences of this was that Saudi Arabia introduced a clause into
their sales contracts that would only allow for the crude oil to be refined by the original
purchaser. Transfer of ownership is prohibited (without the express permission of the
original seller) and as a result there is no actively traded market for Saudi crude oil. Contracts for the supply of Saudi crude oil tend to be ongoing (‘evergreen’) and buyers will
nominate to receive an amount. The Saudis will then decide on how much to deliver.
6.7.5
Pricing sources
When pricing crude oil contracts, participants need to be aware of the current prices.
There are several price reporting agencies that specialise in compiling and publishing
reference prices based mainly on physical market activity, with S&P Platts and Argus
Media being the mostly widely referenced. Depending on the chosen benchmark, prices
may be taken from exchange traded futures such as those published on the CME Group
(WTI) or the ICE (Brent).
6.7.6
Pricing methods
There are several different approaches to establishing the price of a crude contract as
follows:
Official selling prices – this was a traditional method of pricing oil where governments would announce the price for the sale of crude for a fixed period such as one
year, typically to long-term customers.
Fixed price – here the price to be applied has been bilaterally negotiated at a given
level. Anecdotally, it is believed that about 90–95% of crude oil sold is on a long-term
fixed price basis and so therefore the spot market makes up no more than 5–10% of
the market.
Floating price – this uses a price quoted by companies such as S&P Platts or
Petroleum Argus, which is not fixed in advance. The process to determine the
actual price is specified in the contract and may be the average of prices covering a
short period around the delivery date. Sometimes referred as a ‘floating forward’ – a
forward contract whose price at the settlement date will be determined based on
a series of previous traded prices (e.g. the average of daily prices over the previous
month).
Differential to a marker crude – this is where a particular crude oil is quoted as a
differential to one of the marker crudes noted earlier. The differential could move
according to a variety of market factors and could either be positive or negative. For
example, a crude oil that produces a higher yield of gasoline would be in greater
demand during the summer months and so the differential may increase. Equally
194
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
a crude oil that yields a higher proportion of heating oil may be in greater demand
in the winter months, again affecting the size of the differential.
Futures price – this might be relative to a futures price on a particular date or an
average of prices over a previously agreed period. It should, however, be noted that
a futures market does not exist for every type of crude.
Exchange for physicals – this pricing method allows two counterparties to a physical
trade to fix a price using the futures market. A simple numerical example of this
approach is outlined in Section 6.9.
6.7.7
Pricing a cargo of crude oil
When pricing a cargo of crude oil, the contract between buyer and seller will need to
specify certain items:
▪ The underlying benchmark price and its source.
▪ Any agreed differential to the benchmark price.
▪ The pricing period.
▪ Whether the price is FOB or CIF (or another incoterm).
To illustrate the concept, consider the following stylised example:
‘S&P Platts Dated Brent + USD 1.00/bbl., FOB, pricing 2-1-2’
The benchmark is the price for Dated Brent (see Section 6.8) as published by the
price reporting agency S&P Platts plus a differential of USD 1.00/bbl. The oil will be
delivered on a Free on Board (FOB) basis. The last part of the pricing clause requires
a little more explanation. This example illustrates a floating price example, where the
price to be applied to the cargo will be the unweighted average of the published Dated
Brent price around a five-day period (2-1-2). A key part of this formula revolves around
the Bill of Lading date. A Bill of Lading is part of the shipping documentation, which
fulfills three main roles:
▪ It acts as a receipt for the cargo being carried by a vessel.
▪ It is a contract for the carriage of goods.
▪ It is a document of title to the goods being carried.
So, in our example, the pricing period covers the two days prior to, the date of and
the two days following the signing of the Bill of Lading. This may not coincide exactly
with the dates that the crude oil is loaded.
One way to think about such a pricing period is that from the buyer’s point of view
their exposure to the price of crude oil is increasing every day. So, if they agreed to a
cargo of 600,000 barrels and a five-day pricing period then every day their exposure to
the crude oil price increases by 120,000 barrels.
The pricing process in this example is merely illustrative and may vary; participants
may agree, say, the five days after loading or some form of ‘average of month price’.
195
Crude Oil
6.8
TRADING CRUDE OIL AND REFINED PRODUCTS
6.8.1
Overview
There are many crude oil markets, and each has its own quoting conventions, contract
types, and liquidity issues. There are however, arguably, two key global physical markets, namely those for North Sea Oil and West Texas Intermediate. The details of these
markets are considered in Section 6.8 while this section considers some generic trading
principles.
Why are crude oil and the refined products traded?
At a very simple level, trading opportunities can arise for a variety of reasons:
▪ A producer has crude oil available to sell on the open market.
▪ Demand exists within the supply chain (e.g. at the refinery level) for a particular
crude oil.
▪ A crude oil with a particular quality is in demand.
▪ A trader has identified the possibility of making money.
According to their 2017 annual report, on average BP
▪ Produced 1.2 m barrels of crude oil.
▪ Refined 1.7 m barrels of crude oil.
▪ Sold 2.7 m barrels of refined products to retail or commercial customers.
To balance this demand and supply, BP will buy and sell both crude oil and the
refined products from other wholesale sources on an ongoing basis. The annual report
notes that during the period the company traded 2.6 m barrels of crude oil and 3.1 m
barrels of refined products. This would suggest that what they produced was not necessarily refined, but may have been sold to other companies. The annual report also points
out that the company would buy crude oil from other suppliers to optimise their refinery output. This provides the first definition of trading: buying and selling to balance
supply and demand.
Who participates in trading?
Typical market participants include:
▪ National oil companies.
▪ Producers.
▪ Refiners.
▪ Integrated oil companies.
▪ Financial institutions with the capacity to trade the physical products.
▪ Trading houses – for example, in 2017 Vitol traded or shipped around 3.6 million
barrels of oil.
196
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
What risks arise from buying and selling in the physical markets?
Over time a crude oil trader will end up with a portfolio of crude oils. The portfolio will
be comprised of a variety of different exposures:
▪ Crudes priced on different bases (e.g. fixed; floating; futures prices).
▪ Different grades of crude that may be priced off a differential to a certain market
crude.
▪ Crudes that will be bought/delivered on different dates.
▪ Crudes that need to be bought/delivered in different locations.
This gives a second perspective on trading: the mitigation or transformation of the
market risks that arise in the process of buying and selling crude oil. These risks can be
managed by using derivatives.
View driven trading activities
The third and final trading perspective relates to market participants that may wish to
express a view on price movements without necessarily having an underlying economic
exposure to manage. Here the motive is to make money from an anticipated move in
some market variable. These strategies could be implemented using exchange traded
futures or options.
Possible strategies include taking views as follows:
▪ Direction of the price – where the trader buys the commodity if the price is expected
to rise and sells it when the price is expected to fall.
▪ Grade differentials – here the expectation is that the differential to a marker crude
will change.
▪ Shape of the forward curve – when the trader expects a change in the demand and
supply fundamentals at different maturities. These are sometimes referred to as
time spreads. Examples might include:
▪ Trading one quarter or season against another (e.g. Q1 vs. Q; winter vs. summer) to capture some form of seasonality. Some futures contracts will allow the
trader to bundle together individual contracts to form a single longer period or
the trader may have to do this manually by buying or selling a ‘strip’ (e.g. a series
of consecutive contracts).
▪ Trading one calendar year against another.
▪ Freight rates – when rates change due to a sudden change in demand at a certain
geographical location.
▪ Relationships between different crude oil markets – where the trader takes a view
on a particular price relationship that might exist between two markets (regional
spreads). One example of this is the price of Brent crude compared with the price of
WTI. Theoretically, WTI should trade at a premium to Brent but this relationship
can change for several reasons such as:
▪ Declining North Sea output in a period of growing demand.
▪ Maintenance of oil rigs in the North Sea that might lead to a temporary reduction
in output.
Crude Oil
197
▪ High levels of US stocks exerting downward pressure on WTI prices.
▪ Disruption to the supply of other crudes that are considered substitutes for Brent
(e.g. Bonny Light in Nigeria).
▪ The relationship between crude oil differentials, e.g. WTI vs. WT Sour or Bonny Light
vs. Brent.
▪ The crack spread relationship – this is the difference between the price of a finished
product such as gasoline or heating oil and the price of crude oil. There are several
variations of this trade and one example is the gas oil ‘crack’. The price of gas oil
can be broken down into a Brent component and a gas oil crack:
Gas oil price = Brent price + gas oil crack
If gas oil is trading at USD 72.00/bbl. and Brent is trading at USD 60.00, the gas
oil crack would be USD 12.00/bbl. In very simple terms, the gas oil crack spread
presented here could be interpreted as the refiner’s profitability from taking a barrel
of crude and converting it into a higher-value refined product. This gas oil crack is
not a fixed number and will change depending on the relative demand and supply
of the two constituents.
▪ Relationships between the same product traded in different areas – e.g. gasoline
traded in different areas within the same country (Gulf Coast vs. NY Harbor) or
perhaps fuel oil in different continents (e.g. Europe vs. USA).
▪ Product spreads – typical price spreads that might be traded include gasoline and
heating oil or perhaps interfuel spreads such as natural gas against Ultra Low Sulfur
Diesel (ULSD).
▪ Price volatility – option traders will execute option strategies to exploit the perceived
riskiness (i.e. the implied volatility) of the crude oil market. In a similar vein, certain
option strategies are designed to profit from current volatility of the market, such
as gamma trades.
How are oil products traded?
Trading of crude oil and the refined products can be undertaken by a variety of mechanisms. There may be a physical trade, an over-the-counter trade forward or an exchange
traded future. The forwards and futures trades will require physical settlement unless
they have been set up to be cash-settled or closed out prior to expiry. It is possible to
buy options for physical delivery and swaps are also used extensively in the market,
but these will always be cash settled. Option trades may be used as a mechanism to
obtain or dispose of supplies but also as a mechanism to exploit the magnitude of price
movements.
Arbitrage: a question of definition?
‘Arbitrage’ is a word that is often misused in many markets, although in the end its
interpretation is just a question of definition. True arbitrage is the opportunity to buy
a commodity at a certain price and sell it in the same or another market at a higher
price to make a risk-free profit. It is sometimes confused with a spread trade, where
the motive is to exploit an expected change in the relationship between two related, but
ultimately different, commodities.
198
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
A possible true arbitrage for crude oil was illustrated in Chapter 2 and would
involve:
▪ Buying a particular oil in one geographical market.
▪ Knowing what crude products the grade could produce and the total associated
revenues. This would establish the value of the crude to a buyer in the delivery
location.
▪ Adding the cost of freight to the destination.
▪ Adding any associated costs (e.g. insurance costs, costs of financing cargo while in
transit).
▪ Ensuring the income from the final sale at the location of delivery is greater than
all the associated purchase costs.
6.8.2
The Brent complex
At first glance it would appear the phrase ‘Brent Crude Oil’ is referencing a single crude
oil produced in the North Sea. However, nothing could be further from the truth. This is
because the market for Brent (sometimes referred to as the ‘Brent complex’) comprises
of many different components with varying degrees of complexity. It is not helped by
the fact that much of the terminology associated with the market can be confusing.
North Sea crude oil is made up of a variety of different grades, which include:
▪ Brent Blend
▪ Forties
▪ Oseberg
▪ Ekofisk
▪ Troll
▪ Statfjord
▪ Flotta
For many years, the representative price of North Sea Oil was Brent Blend, which
itself is a mixture of crude oils from two different systems (Brent and Ninian) and several
different fields therein. As a result of a decline in North Sea production, the definition
of the Brent crude oil price has widened to include activity in a total of five crudes that
are, broadly speaking, chemically similar: Brent Blend, Forties, Oseberg, and Ekofist
and Troll. These crudes are collectively referred to by the acronym BFOET.
There are a variety of contracts available to trade North Sea oil and they include:
▪ Dated Brent
▪ Brent Forwards
▪ Contracts for Difference
▪ Brent Futures
▪ Exchange for Physicals
6.8.2.1
Dated Brent
The first component of the North Sea market is Dated Brent, which is also sometimes
referred to as ‘North Sea Dated’. As was pointed in the previous section, this is not
the price of a grade of crude oil but a benchmark value that is determined through a
Crude Oil
199
published methodology. As such, it does not have an API density or sulfur specification. Reference to either of these values is usually referencing one of the constituents,
typically Brent Blend. The methodology used by the price reporting agencies (PRAs)
means that the price assessment is based on the lowest price of the five constituent
crudes. However, one of the problems witnessed by the market is that the widening of
the basket has resulted in a significant variation in the quality of crudes, which meant
that to fulfill a commercial contract, producers would often deliver the cheapest of
the five crudes. This oil was termed the ‘cheapest to deliver’. The PRAs have therefore
introduced a process that attempts to adjust for these quality differentials to make a
participant indifferent as to which oil is delivered.
One of the reasons for the continued popularity of Brent as market crude is the fact
that is sometimes considered a spot contract and allows for the pricing of short-dated
crude oil cargoes. However, the term ‘spot’ in relation to the crude oil market warrants
further description, as the term would normally be associated with immediate delivery.
For Brent, normal market practice rarely sees cargoes bought or sold for delivery in less
than 10 days. Dated Brent is a price, which is published daily, and the definition used by
S&P Platts gives an insight into how this part of the market trades. Their price assessment ‘reflects the value of physical crude oil loading 10 days forward from the date of
publication to one full month ahead. The assessed date range will typically stretch to
the equivalent dates of the following month’. So, a price assessment made on 2 January
would cover transactions from 12 January–2 February; an assessment made on 15 January would cover transactions from 25 January–15 February. The Dated Brent price for
any single day is therefore based on volume-weighted prices for transactions encompassing an approximate6 21-day period starting in 10 days’ time. The price quoted for
Dated Brent is a rolling price assessment so that on each day the activity period will
move forward by one day.
It should be noted, however, that the above example describes how a price for Dated
Brent would be assessed for publication by a PRA such as S&P Global Platts. The pricing
of a physical cargo that references Dated Brent as the benchmark would be based on an
average of pre-specified published Dated Brent prices around the time of loading. So if
the agreed loading period were a three-day window, then the cargo would take its value
from an average of the three Dated Brent prices quoted by the designated PRA during
this time.
The respective delivery points for the five crudes in the BFOET basket is shown in
Figure 6.18. The terminal operators will notify producers in advance as to when they
will be able to take possession of their crude oil by means of a tanker over a specified
three-day window. Once a specific three-day loading window has been agreed to then
this cargo will be classified as Dated Brent or a ‘wet’ cargo. In February 2021, S&P Global
Platts announced it would include US WTI Midland crude in its Dated Brent benchmark, with the change becoming effective in July 2022. This was designed to reflect
the decline in output from North Sea crude oils as well as an increase in imports of
US crude oil. At the same time, the price assessment process would change to a Cost,
Insurance and Freight (CIF) Rotterdam basis, meaning crudes will be assessed on the
basis of deliveries into Europe’s largest trading hub, rather than including loadings at
terminals around the region.
6
The exact number of days will depend on weekends and public holidays.
200
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
FIGURE 6.18 Delivery locations for BFOET crude oils.
6.8.2.2
Brent forwards
In the Dated Brent scenario, if a producer is integrated along the supply chain, they may
wish to pick up their crude oil and transport it directly to a particular refinery. However,
a producer knows that since they will be able to load crude oil from the terminal during
any future month, it is possible for them to forward sell a cargo to a buyer who may be
looking to secure their supply. In this situation, it is not possible initially to agree to a
specific loading date with the counterparty and so only the month of delivery is agreed.
The two parties will subsequently agree to the loading dates once the terminal operator
has announced them. This type of forward transaction is often referred to as a ‘paper’
transaction given the lack of a specific loading date.
So, Brent forward contracts lock in the price at the time of contracting of a commodity for delivery in a specific future month beyond the Dated Brent assessment period.
Somewhat confusingly, forward contracts in Brent may be referred to in several
different ways as:
▪ Brent forwards
▪ Paper Brent
▪ Cash BFOET
▪ Paper BFOET
▪ Cash forwards
For simplicity, these types of contracts will be referred to in the text as forwards.
The price quotation for this type of forward represents the value of a cargo for physical
delivery within the month specified by the contract.
At the time of writing, for a given day, S&P Global Platts will quote a value for Dated
Brent as well as three Brent forwards. If it is early March, they will quote forward price
201
Crude Oil
for May, June, and July. These three months will be quoted for the remainder of March
when the contract months will ‘roll’. So, from April, the forwards quoted would be June,
July, and August.
The way in which this paper forward turns into a wet Dated Brent contract is outlined in standard market documentation referred to as ‘SUKO 90 terms’ (Shell UK Oil
Company). The terms state:
‘Seller shall declare to the buyer the grade deliverable, the laydays and the cargo
reference number in respect of the cargo between 0900 and 1600 hours (London time) . . . not later than on the last day of nomination . . . The last day of
nomination shall be one full month prior to the first day of the laydays . . . ’
Assume that on 7 March of a given year, a Brent forward is executed for delivery
in May. In the month prior to delivery (April in this case) the managers of the loading
installation will list the dates in May when the contract seller can load oil. This set of
dates is referred to as the ‘laycan’, which is the three-day window where the buyer can
arrange for a vessel to arrive for loading. For example, the assigned loading dates may
be 1–3, 10–12, 20–22 and 26–28. In the Brent Forward market, the contract seller has
the right to nominate any of the loading dates to another buyer so long as they give the
requisite one-month notice.
If on 17 April the seller wished to nominate a particular set of delivery dates to physically settle the forward, they would no longer be able to use the first two sets (1–3 May
and 10–12 May) due to the requirement to give one month notice. Cargos for delivery
on those earlier dates would be classified and traded as Dated Brent.
If the parties to an agreed trade do not wish the deal to go to physical delivery, it
will have to be cash settled or ‘booked out’ to use the industry jargon.
Table 6.7 illustrates the main characteristics of the Brent forward contract, which
although OTC in nature, does have an element of standardisation within the market.
Reference may sometimes be made to partial contracts, which describe a forward contract with‘standard’ terms apart from the fact that the trade size will be less than the
600,000 barrels.
6.8.2.3
Contracts for Difference
Brent Contracts for Difference (CFD) are one period swaps that give a user exposure to
the differential in price between Dated Brent and a Brent Forward Contract. However,
this differential is not the difference between the current Dated Brent value and the
first available Forward Brent Contract. This type of transaction within exchange traded
futures markets would be referred to as the ‘dated to frontline’ contract.
CFD quotes represent the difference (either positive or negative) between the current third month7 Brent forward quote and Dated Brent for a stated future period, in
USD/bbl. Therefore, the CFD quote represents the value of time between the maturities of the two contracts. The quote for a given period will change as the relative demand
and supply for the two constituent components evolves.
7
Confusingly, S&P Global Platts describe this as the ‘M2’ contract.
202
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The relationship could therefore be expressed as follows:
CFD quote = Forward Dated Brent minus Third Month Brent Forward
However, since both the CFD and the Brent Forward are observable it is perhaps
more useful to rearrange the formula to derive an implied Dated Brent value for a forward period. Rearranging the formula gives:
Forward Dated Brent = CFD plus Third Month Brent Forward
Another interpretation of the equation is that buying a CFD and taking a long position in the third forward month will lock in the purchase price for Dated Brent at a
certain time in the future.
Say it is early March and the following bid and offer prices are being quoted in the
Brent market:
Dated Brent
Brent forwards
May Brent
June Brent
July Brent
Contracts for difference
CFD 1-week 06–10 March
CFD 2-week 13–17 March
CFD 3-week 20–24 March
USD 65.00–65.02
USD 67.90–67.92
USD 68.70–68.72
USD 69.49–69.51
−2.01/−1.99
−1.60/−1.58
−1.41/−1.39
CFD 8 − week 24 − 28 April − 0.35∕ − 0.33
Diagrammatically, this is shown in Figure 6.19. For brevity purposes only the offer
prices are shown.
The normal market size for Brent CFDs is between 50,000 to 100,000 bbl. and they
are traded with weekly maturities out to eight weeks. They can also trade for bi-monthly
and monthly periods. Like any contract for difference, the counterparties agree to a fixed
differential at the point of the trade and then subsequently settle on the actual differential (sometimes referred to as the ‘floating’ rate) at the contract maturity. Settlement at
Dated
Brent
Contracts for difference
(CFDs)
Brent forwards
Mar
3
Mar
06–10
Mar
13–17
Mar
20–24
Mar
27–31
Apr
03–07
Apr
10–14
Apr
17–21
Apr
24–28
May
forward
June
forward
July
forward
65.02
-1.99
-1.58
-1.39
-1.17
-0.97
-0.68
-0.51
-0.33
67.92
68.72
69.51
FIGURE 6.19 The relationship between, Dated Brent, Contracts for Difference and Brent
forwards.
203
Crude Oil
TABLE 6.5 Interpreting a CFD quote.
Market maker
Market user
Bid
Offer
Buy
Pay quoted differential (fixed)
Receive actual differential (floating)
Profits if actual differential is more
positive/less negative than quoted
differential
Sell
Receive quoted differential (fixed)
Pay quoted differential (floating)
Profits if actual differential is less
positive/more negative than
quoted differential
Sell
Receive quoted differential (fixed)
Pay quoted differential (floating)
Profits if actual differential is less
positive/more negative than quoted
differential
Buy
Pay quoted differential (fixed)
Receive actual differential (floating)
Profits if actual differential is more
positive/less negative than quoted
differential
maturity is based on the average of the daily differential between Dated Brent and the
original Brent Forward contract.
CFD are quoted on a bid/offer basis (see Table 6.5) so when the market is in contango, the quote is negative and the CFD bid price is greater than the offer. This ensures
that when the implied forward price for dated Brent is derived it will follow the normal
convention of low bid, high offer. A market in backwardation will result in a positive
CFD quote with the bid price lower than the offer.
The fact that the CFD quote may be negative combined with the use of ambiguous
phrases such as ‘spread widening’ and ‘narrowing’ can make the interpretation of CFD
quotes confusing.
To illustrate the principles, consider an example of a CFD quote of +USD 1.50/+
USD 1.55. If a trader believed that the differential would increase (i.e. become more
positive), then they would buy the CFD at the offer side of the market as a market user.
If at settlement the actual differential had increased to USD 1.75, a payment of USD
1.55 would be made and a payment of USD 1.75 received, representing a USD 0.20 net
receipt per barrel. If it were assumed that the deal was executed on a notional basis of
50,000 barrels, this would equate to a cash settlement of USD 10, 000 (50,000 barrels x
USD 0.20)
Now consider that the quote is −USD 2.10/−USD 2.05. Once again it will be
assumed that the trader believes that differential will increase (i.e. become less
negative) and so decides to buy the CFD at the offer side of the market. If at settlement
the actual differential had increased to −USD 1.85 the terms of the contract necessitate
making a payment of −USD 2.05 and receiving a payment of −USD 1.85. But a
payment of a negative sum is a receipt as two negatives make a positive! A receipt of
a negative sum would therefore represent a payment and the trader would be a receiver
of USD 0.20 a barrel, which makes the two examples consistent.
The previous CFD example focused on the use of the instrument to exploit how
the relative prices of Dated Brent and Brent Forward were expected to evolve. CFD can
also be used as a mechanism to hedge a position in Dated Brent. Let us assume that it
204
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 6.6 Prices used to settle a Contract for Difference.
Date
Dated
Brent
(USD)
June
forward
(USD)
Differential of Dated
Brent minus June
forward (USD)
Monday, 20 March
Tuesday, 21 March
Wednesday, 22 March
Thursday, 23 March
Friday, 24 March
69.48
69.49
69.50
69.51
69.52
70.79
70.78
70.80
70.82
70.81
−1.31
−1.29
−1.30
−1.31
−1.29
is early March and a trader has agreed to buy a cargo of crude that will be priced on
a floating basis using Dated Brent. Since he has not yet taken delivery of the crude, in
market jargon he is said to ‘short crude for forward delivery’. The cargo will be loaded
on 21–23 March and the invoice price will be based on an average of Dated Brent prices
centered on the loading date. The trader looks at the price screen and notes that the
current value for Dated Brent is USD 65.02. He is concerned that the price of Dated
Brent will rise and so decides to hedge his exposure. He could use the May forward to
hedge the risk but runs a risk that the price of the forward may not track the movements
in Dated Brent used to price his physical contract (‘basis risk’).
To hedge this exposure the trader will have to do the following:
▪ Buy a CFD that covers the week in which the physical cargo will be priced and,
▪ Take out a long position in the third month Brent Forward contract from which the
CFD is priced.
Using the figures quoted earlier the trader’s position using market user rates
would be:
▪ Long one cargo of BFOET loading 21–23 March; priced at an average of Dated Brent
around the time of loading.
▪ Buy a CFD for the week 20–24 March; priced at −USD 1.39. Since the price is a
negative, they will receive this value and pay the realised differential at settlement.
▪ Long one Brent Forward for June delivery at USD 68.72.
Using the relationships established earlier it is possible to back out an implied forward value for Dated Brent for the delivery week of 20–24 March.
Since mathematically the CFD quote is the Forward Dated Brent value minus the
second month Brent Forward, we can back out the Forward Dated Brent price.
−USD 1.39 = Implied Forward Dated Brent − USD 68.72
Implied Forward Dated Brent = 68.72 − 1.39 = 67.33
Buying the CFD and going long the third month forward is now economically
equivalent to buying Dated Brent on a forward basis at a price of USD 67.33. Assume
that over the course of the loading week (i.e. the week ending 24 March), the prices in
Table 6.6 are observed.
205
Crude Oil
The result of the hedge will be:
▪ The physical cargo will price as the average of the Dated Brent contracts that were
observed on 21–23 March. This returns a value of USD 69.50.
▪ The average of the Dated Brent to June forward differentials observed over the week
is used to settle the CFD week. The average is calculated as −USD 1.30. The trader
receives USD 0.09/bbl.
▪ The trader closes out their June forward position. If this is assumed to be done at
the Friday, 24 March price shown above, then they will show a profit of USD 2.09
▪ The total cost per barrel is therefore USD 69.50 − USD 0.09 − USD 2.09 = USD
67.32. This price is within USD 0.01 of the implied Forward Dated Brent value calculated earlier. The difference is attributable to the fact that the CFD is calculated
as the average of five values, but the June forward is closed out using a single value.
This may seem somewhat complex and so in order to make the hedging of the physical exposure easier, the purchaser may opt to go for a futures based pricing agreement,
which could then be hedged by purchasing the same futures contract.
6.8.2.4
Brent futures
There is also an active exchange traded futures market that allows for crude oil to be
traded for delivery in a specified future month. As with any futures contract, it will have
standardised terms and conditions. The main Brent contract is traded on the Intercontinental Exchange (ICE). According to the ICE, ‘The ICE Brent Crude futures contract
is a deliverable contract based on EFP delivery with an option to cash settle against the
ICE Brent Index price for the last trading day of the futures contract’.
Trading in the Brent futures contracts ceases at the close of business on the last
business day of the second month preceding the relevant contract month. So, the March
contract will expire on the last business day of January. Contracts are available for maturities out to 96 months but the longer-dated contracts may offer different degrees of
liquidity. One of the characteristics of every futures contract is that the price of the future
should eventually converge to that of the underlying. So, upon the expiry of the futures
contract, its value should be the same as the asset to which it is referenced. If this were
not the case, then in theory it should be possible to buy and sell the physical and the
future and make a risk-free profit. The futures contract settles against the ICE Futures
Brent Index:
‘The cash settlement price for the ICE Brent Future is based on the ICE Brent
Index (‘The Index’) on expiry day for the relevant ICE Brent Futures contract
month. The Index represents the average price of trading in the BFOET
(Brent-Forties-Oseberg-Ekofisk-Troll) cash or forward (‘BFOET Cash’) market
in the relevant delivery month as reported and confirmed by the industry
media. Only published cargo size (600,000 barrels) trades and assessments are
taken into consideration in the calculation”.
206
6.8.2.5
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Exchange for Physicals
The specification of a crude oil futures contract can vary between exchanges. Some of
the contracts will involve physical delivery while others may cash settle; some contracts
may offer an element of both. Additionally, only a small proportion of all futures contracts will go to final delivery for a variety of reasons:
▪ The contract maturity may not match the timings of the underlying exposure.
▪ The type of crude oil referenced by the future may differ from the underlying exposure.
▪ The location of any physical delivery may differ from the underlying exposure.
▪ If the contract is physically delivered, the exchange’s settlement algorithm may
select an unfamiliar entity or indeed multiple entities with which they must deal.
Exchange for Physicals (EFP) is an instrument that helps manage these types of
issues. As the name suggests there is an exchange of a physical holding of crude
oil against an equivalent futures position. EFPs are quoted as a differential, which
attempts to capture three main elements:
▪ Time – since the two holdings have different times to delivery (immediate against
some future date), then it is unlikely that they will have the same value. For
example, the shape of the forward curve is likely to have some impact with the
parties wishing to consider whether the market is in backwardation or contango.
▪ Quality – participants can also agree to a transaction where the physical crude oil
differs from that referenced by the future so a quality differential will need to be
agreed.
▪ Location – they can also negotiate the actual location where the physical crude oil
will be delivered, which again could differ from that outlined in the future.
EFPs are quoted as a differential to a particular futures maturity, which can be
agreed between the two participants. The EFP does not necessarily need to reference
the futures contract with the earliest maturity. So, if the agreed future were trading at
USD 62.00/bbl. and the EFP was quoted at +USD 0.10 then the price of the physical to
reflect time, quality, and location would suggest a value of USD 62.10.
The following example illustrates the principles. It is early February, and a producer
has one million unsold barrels of crude oil. An oil refiner is long 1,000 futures (where
each future references 1,000 barrels) as they have a wish to protect themselves against
rising prices. It is assumed that the future was originally entered into at a price of USD
63.00/bbl. The two parties enter an EFP (probably via a broker) whereby the producer
will agree to sell the physical crude oil to the refiner and buy an equivalent futures
position. The refiner will take delivery of the crude oil and agrees to sell their existing
futures position. Both parties agree to reference the EFP to the shortest-dated futures
contract, which is now assumed to be trading at USD 62.00. They agree to a differential
of +USD 0.10 to reflect the difference in crude quality, the location where the crude will
be delivered to, and the appropriate date. Both parties will then be obliged to register
the EFP with the relevant exchange.
207
Crude Oil
From the producer’s perspective, they will sell their physical crude oil at a price
of USD 62.10, but since they now have a long futures position initiated at USD 62.00,
the final value of the commodity will only be determined when this position is closed.
They are at liberty to close this position whenever they want and may choose to do so
over a period. However, for ease of illustration let us suppose it is closed immediately
at the same futures price used to settle the EFP. They will make no profit or loss on the
futures position since it was opened and closed at the same price, so the final value of
their physical sale is USD 62.10.
From the refiner’s perspective, they have:
▪ Agreed to purchase physical crude oil at USD 62.10.
▪ Closed out a long futures position established at USD 63.00 at a price of USD 62.00
to realize a USD 1.00/bbl. loss.
▪ Therefore, the cost of their physical purchase was USD 62.10 + USD 1.00 = USD
63.10.
Like the producer, the refiner is not obliged to close out the entire futures position
when entering the EFP. For example, many of their refined products will be sold on
a rateable basis throughout a month and so the price they receive for these may be
calculated as some form of average. It would seem logical that the refiner may also wish
to price their crude oil expense on a similar basis, and so may decide to unwind their
futures hedge on a piecemeal basis over a similar time horizon.
6.8.2.6
Summary
Table 6.7 compares the main features of the three main types of Brent contract.
6.8.3
US crude oil markets
The main type of crude oil traded in the USA is West Texas Intermediate (WTI). This
is usually quoted for delivery at Cushing, Oklahoma, although S&P Global Platts does
provide quotes for alternative delivery locations such as Midland, Texas. There is a range
of other crude oil markets with different delivery locations that include:
▪ West Texas Sour (delivered into Midland Texas).
▪ Light Louisiana Sweet (St. James, Louisiana).
▪ Mars (Clovelly, Louisiana).
▪ Alaska North Slope (Long Beach, California).
Physical WTI price quotations are driven by the pipeline companies’ requirement
that all deliveries for a given month be notified by the 25th of the previous month. So, for
the period of 26 March–25 April the quoted price on any single day is a representation
of crude oil transactions to be delivered during the month of May. On 26 April, the oil
price will be reflective of transactions to be delivered during June.
Another price index worth mentioning is the ‘American Sour Crude Index’
or (ASCI). This is published daily by the price reporting agency Argus Media and
208
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 6.7 Summary of Brent contracts.
Exchange or
OTC
Underlying
asset
Pricing point
Exposure
period
Loading date
Standard size
(bbl.)
Settlement
Dated Brent
Brent forward
Brent futures
OTC
OTC
Exchange
Brent, Forties,
Oseberg, Ekofisk,
Troll
FOB – various
locations in North
West Europe
Spot price of crude
referencing agreed
transactions 10
days – one-month
post publication
Specific date agreed
in contract
Brent, Forties,
Oseberg, Ekofisk,
Troll
FOB – various
locations in North
West Europe
1st–31st of the quoted
calendar month;
greatest liquidity in
near dated
contracts
Nominated by seller
at least one month
prior to delivery
Brent, Forties, Oseberg,
Ekofisk, Troll
600,000
600,000
Physical settlement by
means of EFP with an
option to cash settle
against Brent Index
1,000
Physical
Physical/cash
Physical/cash
FOB – various locations
in North West Europe
Consecutive calendar
months out to about 8
years
represents the price of medium sour crude on the US Gulf Coast. This became popular
when Saudi Arabia decided to reference their US sales of crude away from WTI
to an index that they felt more closely resembled the product they were looking to
sell. The index is now also used by Kuwait and Iraq when selling into the USA. The
component crude oils are Mars, Poseidon, and Southern Green Canyon.
Futures on crude oil and several related petroleum products are offered on the CME
Group exchange, with maturities extending to 10 years. Their light, sweet, crude oil
contract is physically deliverable by pipeline at Cushing, with each lot consisting of
1,000 barrels (42,000 US gallons). The contracts expire on the third business day prior
to the 25th calendar day of the month preceding the delivery date. This allows the expiry
of the future to dovetail into the settlement conventions of the physical market noted
above. So, if a participant were trading a future with May delivery this contract would
expire on 22 April, allowing any short that decides to go to final settlement three days
to make the necessary arrangements to schedule the delivery.
Despite allowing for physical settlement most futures positions are closed out
before contract expiry, which means that buyers sell out of their positions at prevailing
market prices whilst sellers do the opposite. Closing out of positions occurs even
in situations where market participants wish to buy or sell physical quantities. This
is because the standardised terms of futures delivery are usually too restrictive for
physical market participants. The CME Group also offers contracts on several refined
products, some of which are summarised in Table 6.8.
209
Crude Oil
TABLE 6.8 Summary of CME contract specifications.
Contract Size
Quote
Minimum
price
fluctuation
Traded
periods
Delivery/
Settlement
Delivery
location
Brent
Crude
NY Harbor
Ultra Low
Sulfur Diesel
(ULSD)
RBOB
Gasoline
1,000 barrels
USD and cents
per barrel
0.01 USD per
barrel
1,000 barrels
USD and cents
per barrel
0.01 USD per
barrel
42,000 gallons
USD and cents
per gallon
0.0001 USD
per gallon
42,000 gallons
USD and cents
per gallon
0.0001 USD per
gallon
Current year
and the next
10 years
Physical
Current year
and the next
seven years
Cash
Three years
Three years
Physical
Physical
Cushing, OK
N.A.
New York Harbor
New York Harbor
West Texas
Intermediate
(WTI)
Source: Based on CME Group
NOTES
▪ At the time of writing, the CME also offered a cash settled WTI contract.
▪ The ULSD contract may be casually referred to as a heating oil (HO) contract. Heating oil is diesel fuel although it may contain a coloured dye to indicate it cannot be
used in vehicles for tax reasons.
▪ RBOB stands for Reformulated Gasoline Blendstock for Oxygen Blending. To sell a
refined product, it has to meet a series of technical and legal/environmental specifications. Intuitively, RBOB can be thought of as a common building block used in
the process of producing gasoline. It is analogous to base metals futures, where the
contract specification is an intermediate rather than a final product.
6.8.3.1
Calculating the price of US physical crude oil
Although the reader may feel the inclusion of this topic after a discussion of the futures
market is somewhat odd, the pricing of physical crude oil is significantly impacted by
futures prices.
Suppose a producer is considering selling physical WTI crude oil at Cushing during
the month of May. He and the consumer agree to use the conventional pricing method
of:
NYMEX calendar month average (CMA) plus the NYMEX roll
Figure 6.20 helps illustrate the concept and the associated timelines.
The CMA component means that for the delivery month of May, part of the agreed
payment will consist of the arithmetic average of the front month futures settlement
210
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
NYMEX roll
Calendar month average
Daily calculation over the period:
Arithmetic average of
front month futures settlement
price for delivery month
((M1 futures price – M2 futures price) x 0.6667)
+
((M1 futures price – M3 futures price) x 0.3333)
The roll is the arithmetic average of the daily calculations
Individual trading days
25th
April
1st
May
Individual trading days
31st
May
FIGURE 6.20 Calculating the NYMEX roll and the calendar month average.
prices observed over that month. This means that the final price will not be known
until after the last trading day of the respective delivery month. However, during the
month of May the futures price is initially referencing prices for June delivery before it
expires towards the end of the month, when July maturities becomes the front month
contract.
The NYMEX roll component adjusts this price to make it more representative of
prices for May delivery. Refer again to Figure 6.20. The May futures will be the front
month contract from the period 26 March–25 April.8 For each day of this earlier period,
the following calculation is performed:
((Month One futures price − Month Two futures price) × 0.6667) +
((Month One futures price − Month 3 Three futures price) × 0.3333)
By way of illustration suppose that on the 26th March the following settlement
prices were observed:
May (month 1)
June (month 2)
July (month 3)
40.28
40.58
40.87
The calculation for that day would be:
= ((40.28 − 40.58) × 0.6667) + ((40.58 − 40.87) × 0.3333)
= −0.20 + −0.10 = −0.30
8
Assuming the 25th is a good business day, the May contract will expire, and June will become
the front month.
211
Crude Oil
This value would be calculated for each business day over 26 March–25 April period
with the NYMEX roll being the average of the results.
So if CMA for May is calculated as, say, USD 41.00 and the NYMEX roll was −USD
0.50 then the producer will receive USD 40.50 for each barrel sold, irrespective of when
during the agreed month they may have been delivered.
6.8.3.2
Negative WTI prices
As the manuscript for this book was being prepared, a significant number of countries
introduced radical social and economic measures to prevent the spread of COVID-19.
As a result, the demand for crude oil fell faster and farther than at any point in history
(Economist, 2020). Economic activity slowed considerably, and so more crude oil was
being placed into storage. Since WTI is a landlocked crude oil with a pricing point at
Cushing, Oklahoma, the availability of storage at this location has a significant impact
on the reported price. On 20 April 2020 for the first time in its history, WTI crude oil
for May 2020 physical delivery closed at a negative price. The market opened at USD
17.73/bbl. and traded as low as USD −40.32 to eventually close at USD −37.63, a daily
range of USD 58.05. The May 2020 contract eventually expired the following day at USD
10.01/bbl. Arguably this was because of the structure of the WTI market as the price for
Brent crude oil remained positive.
Section 12.7 illustrates the concept of a short squeeze whereby participants with
short futures positions may end up pushing the price market price higher to avoid having to go to physical delivery. Since the price of crude oil collapsed, then this implies the
opposite; there must have been participants who were long the futures and realised if
they were unable to close out their positions they would have to take physical delivery.
This would mean that they would then be forced to either move the crude oil out of
Cushing or place it into a local storage facility.
At the time, the EIA reported that stocks in Cushing were approximately 60 million
barrels and with the installation’s working capacity estimated at 76 million barrels, so it
was about 79% full. According to figures published by the CME Group, the open interest
(i.e. outstanding contracts) as at the close of 20 April 2020 was 13,044 contracts, equivalent to 13,044,000 barrels. If all these contracts had gone to final delivery, then storage
capacity at Cushing would have been very close to its limit. Indeed, reports (Financial
Times, 2020a), suggested that it was highly likely that much of the apparent storage
availability had been booked to accommodate future inflows. So it seems likely that
rather than be faced with the issue of trying to dispose of oil that could be neither stored
or transported out of the area, many ‘longs’ were willing to sell at any price. Of course,
in the end, a negative oil price means these longs were paying to sell their oil.
There were also suggestions that investor activity may have exacerbated the problem. An investor who uses a future to gain a bullish exposure to crude oil would go long
the contract but would need to ‘roll’ this exposure as the maturity approached to avoid
physical delivery. The rolling process involves selling out of the existing long position
and re-establishing the long position in the next available maturity. If the investor position were relatively large, then the selling pressure to instigate the roll would lead to
further downward pressure on the price.
212
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
6.9
MANAGING PRICE RISK ALONG THE SUPPLY CHAIN
Section 6.3 provided an overview of the crude oil supply chain. Based on that analysis,
it is possible to identify the different risks that are faced by each of the participants.
Producers – exposed to the absolute price of crude as well as an adverse move in
quality differentials. The exposure to the level of crude oil prices is sometimes referred
to as ‘flat’ price risk.
Refiners – on the assumption that they have purchased the oil on an FOB basis,
there are several exposures to consider:
▪ The cost of financing the purchase.
▪ Freight and insurance costs.
▪ Costs of storing the crude oil or refined products.
▪ Refinery margins (‘crack spreads’)
Consumers – they may be exposed to:
▪ The absolute price of the refined product.
▪ A movement in the price differential between the location used to price the commodity and the end location to which it will be delivered.
▪ The fact that a liquid hedge for their product may not exist and so may be forced to
implement a proxy hedge (e.g. hedging jet fuel exposures with crude oil).
Some of these risks will be considered below but for space purposes other topics
will be considered in separate chapters (e.g. financing and freight costs).
6.9.1
Producer Hedges
In the following examples, structures that are unique to crude oil are emphasised. To
avoid repetition, interested readers are referred to the chapter on base metals for generic
vanilla and exotic option strategies.
Exchange traded futures
Although the use of futures would appear to be straightforward, there is the possibility
that an element of basis risk may exist. Basis risk is used to describe several different
situations, which include:
▪ Using futures in one product to hedge an underlying exposure in another (e.g. using
crude oil futures to hedge a jet fuel exposure).
▪ A mismatch between the timing of the exposure and the protection period covered
by the future.
▪ Termination of a futures hedge prior to maturity, where there is an underlying physical exposure. Here the basis risk is that the cash and futures prices have not moved
in parallel resulting in a hedge that is not 100% efficient.
Crude Oil
213
▪ Physical delivery of the commodity in a location that differs from that used to price
the hedging instrument.
▪ Quality differences between the physical exposure and that specified by the hedging
instrument (e.g. hedging a Bonny Light exposure with a Brent Future).
Consider an example of a US oil producer who in early December is looking to
hedge the price risk of his planned production in the first quarter of the following year.
He has agreed to sell 50,000 barrels per month to a refiner in each of the first three
months. The contract will be settled on a floating rate formula with the final agreed
price based on the average of quoted S&P Global Platts prices, published daily in the
month preceding the delivery of the crude oil. So, for the 50,000 barrels to be delivered
on the first business day of January, the price paid by the refiner will be the average of
the S&P Global Platts WTI prices published daily in December. The February delivery
will be based on the average of January prices, and a March settlement will be based on
the average of February prices.
If the producer was concerned about a fall in the price of WTI, then he could sell
three lots of 50 futures contracts to mitigate the risk. In a perfect world each futures contract would have a maturity that matches the underlying physical deliveries. However,
it is not that simple with the WTI contract. His January delivery will be priced at the
average of December’s cash prices and so the final price will not be known January. At
that point in time the January WTI future will have expired. (The WTI contract expires
three business days prior to the 25th calendar day of the month preceding the delivery
day). Hence, he may choose to hedge the December price movements using a February
contract, which will mature towards the end of January.
Because the price exposure of the physical sale and the futures cover different time
periods, the producer runs the risk that the cash and futures prices will not move in
tandem; one example of basis risk. Fortunately for the hedger, the chances of futures and
cash prices decoupling, i.e. the basis risk, is usually less of a concern than the absolute
price risk.
The following example illustrates the concepts. On this occasion we have assumed
that the physical oil contract was priced at a single point in time. On 1 December a producer enters into a long term contract whereby he agrees to sell forward 50,000 barrels
per month of physical WTI crude oil for the first three months of the following year
according to a price equal to the S&P Global Platts assessed price for WTI at Cushing
on the first good business day of each month. The producer fears that cash prices may
decline over the period resulting in lower revenues.
To hedge their exposure to the pricing windows at the start of each month, the
producer decides to initiate a short hedge by selling a strip of futures contracts. Hence
to hedge its January, February, and March sales, it sells February, March, and April
CME WTI futures. On the pricing days in question the prompt month futures position
is bought back at the prevailing market rate to unwind the hedge. Sometime after each
pricing date the producer delivers 50,000 barrels of oil and receives a cash amount based
on the S&P Global Platts assessments in return. Table 6.9 shows the outcome of the
hedging programme and the effective price realised by the producer.
Table 6.9 shows that the producer was able to lock in an average price of 68.28
USD/bbl. for its sale and that this outperformed an un-hedged position which would
214
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 6.9 Producer hedge using WTI futures.
Date
Physical
Futures
Dec 1
Sells 50,000 bbl. per
month on a forward
basis for Jan, Feb,
and March delivery
S&P Global Platts cash
price = 68.37
S&P Global Platts cash
price = 67.93
S&P Global Platts cash
price = 67.55
S&P Global Platts cash
price = 68.79
Sells 50 lots per
month of WTI:
February = 69.67
March = 70.04
April = 70.63
Jan 1
Feb 1
Mar 1
Buy 50 lots
February = 69.32
Buy 50 lots
March = 69.29
Buy 50 lots
April = 71.15
Futures
Gain/Loss
(USD/bbl.)
Effective
Price
(USD/bbl.)
(69.67 − 69.32)
= 0.35
(70.04 − 69.29)
= 0.75
(70.63 − 71.15)
= −0.52
(67.93 + 0.35)
= 68.28
(67.55 + 0.75)
= 68.30
(68.79 − 0.52)
= 68.27
Average S&P Global Platts cash price = (67.93 + 67.55 + 68.79)/3 = 68.09 USD/bbl.
Average effective price = (68.28 + 68.30 + 68.27)/3 = 68.28 USD/bbl.
have yielded a lower average price (USD 69.09) over the same period. On the other
hand, had prices evolved in a positive direction then they would have locked in its sales
revenues at the expense of being able to benefit from a price rise.
It is important to note that the effective sale price of 68.28 USD/bbl. was not quite
equal to the prevailing cash price at the time the hedge was initiated (68.37 USD/bbl.).
This is because cash prices and futures did not always move in tandem during the period
in question.
The basis can be quantified as the spot price minus the future price and its evolution
for this trade is documented in Table 6.10. In all the examples the basis decreased (i.e.
became more negative).
The basis at the closeout can also be used to determine the effective price paid for
the crude oil using the following relationship:
Original futures price plus basis at closeout = effective price of crude
TABLE 6.10 Evolution of the basis.
Initial value for
basis
Value of basis at
close out of
future
February futures
March futures
April futures
68.37 − 69.67 = −1.30
68.37 − 70.04 = −1.67
68.37 − 70.63 = −2.26
67.93 − 69.32 = −1.39
67.55 − 69.29 = −1.74
68.79 − 71.15= −2.36
215
Crude Oil
To illustrate, consider the physical consignment to be delivered in January. The
original February futures price was 69.67 USD and the basis on close out was −1.39
USD (67.93 − 69.32). This would infer an effective sale price for the crude of 68.28 USD,
which indeed was the case:
69.67 − 1.39 = 68.28
In executing this hedge, the producer took on the risk that the cash WTI price would
weaken even further relative to the prompt WTI futures contract. The producer was
willing to take this risk having determined that the basis between cash WTI and futures
was far less volatile than the outright cash price. That is, they would achieve a less
favourable sale price if the basis decreased or became more negative. For example, on
the close out of the February future, the effective sale price would only have been 67.67
USD if the basis had been −2.00 USD.
Hedging crude oil priced over a multi-day loading period
In some cases, the physical supply contract would be based on an average of prices over
a period to reflect the loading period. In this case the futures position would have to be
unwound in equal amounts as the physical contract is priced. So, if the physical contract
is priced as the average of prices over five days, then 1/5 of the futures hedge would have
to be unwound daily.
Suppose it is late May and a producer of crude oil is looking to sell 300,000 barrels
of crude oil for delivery in three months’ time (e.g. late August). They agree to a commercial contract with an oil refiner whereby the cost of the physical crude oil will be
based on the average of the ‘near month’ future’s closing price over the three days it
will take to load the cargo onto the refiner’s vessel. In the market jargon the producer is
‘long’ the market, i.e. they will be adversely impacted by falling prices. As a result, they
decide to sell futures to hedge their exposure. So, at the end of May, the refiner decides
to buy 300 October contracts, as this will be the ‘near month’ contract on the expected
delivery date. The example assumes that each contract is for 1,000 barrels and that the
contracted futures price is USD 59.37.
Suppose that the closing futures prices for the ‘prompt’ contract for each of the three
August delivery dates turn out to be:
Date #1
Date #2
Date #3
USD 62.00
USD 63.50
USD 61.25
Under the terms of their physical contract with the refiner, the producer receives
the average of these three values multiplied by the number of barrels. The average of
these three prices is USD 62.25/bbl. so for the 300,000 barrels this amounts to USD
18,675,000.
216
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The producer decides to unwind 1/3 of his futures position at these closing prices;
that is, he sells 100 futures every day. The resulting profit and loss are as follows:
Date #1:
USD 62.00 − USD 59.37 = USD 2.63 loss
Trader buys 100 futures (i.e.100,000 barrels)
On 100,000 barrels this equals a loss of USD 263,000
Date #2:
USD 63.50 − USD 59.37 = USD 4.13 loss
On 100,000 barrels this equals a loss of USD 413,000
Date #3:
USD 61.25 − USD 59.37 = USD 1.88 loss
On 100,000 barrels this equals a loss of USD 188,000
Total profit and loss on futures position = USD 263,000 + USD 413,000
+ USD 188,000
= USD 864,000 loss
The amount received by the producer considering their futures hedge would be:
Income received from sale of physical crude minus futures loss
= USD 18,675, 000 − USD 864,000
= USD 17,811, 000
On a per barrel basis this is equal to selling crude oil at a price of USD 59.37 – a sum
equal to the original futures price.
Admittedly, since prices increased then with hindsight, the producer would have
been better off doing nothing but the example shows that the futures represents a binding commitment and do not possess the asymmetry of options.
The hedge was 100% efficient, but this is unlikely to be the case:
▪ It is assumed that the underlying crude oil and the future are for the same type of
crude oil (e.g. Brent).
▪ The pricing of the physical contract references a futures contract rather than
another index. This allows the hedge to move in line with the underlying asset
price.
▪ The termination of the futures hedge was done at the closing price, which also
determined the price of the physical contract.
Crude oil swaps
A crude oil swap could be used to transform a specific price risk faced by a client. Let
us say that a producer has a contract to sell oil on an ongoing basis at a fixed price.
Crude Oil
217
The producer believes that the price of crude will rise and so wishes to benefit accordingly. The refiner enters a crude oil swap where they agree to pay a fixed price per barrel
(say USD 50.00) to receive a floating benchmark price such as the average of exchange
traded futures prices. Depending on how the swap is structured the fixed cash flows
should cancel and the net effect is that the producer ends up selling crude oil on a floating price basis.
Although somewhat ambiguous, the market may use the terms ‘buy’ and ‘sell’ in
relation to the swap instrument. Here we will define the purchase of a swap as an
instance where an entity pays a fixed price and receives the floating or variable index.
Selling a swap will be defined as a receipt of fixed against a payment of floating. To avoid
ambiguity, it is perhaps preferable to state who is paying or receiving fixed.
The swap transaction presented here covers an exact calendar month with the fixed
price set at USD 50.00 payable by the producer. The floating index will be based on
a straight average of futures prices published daily by the relevant exchange over the
period of the contract month. At the end of the calculation period let us assume that the
average futures price was USD 49.50 and that the contract was executed on 70,000 bbl.
Just like all swap contracts, when cash flows are timed to coincide a single net payment
is due. The producer would therefore be a payer of USD 0.50 per barrel under the terms
of the swap. When translated into a cash amount (there is no physical delivery of crude
oil under the swap) this would equate to USD 35,000 (70,000 barrels x USD 0.50). The
producer will sell their physical crude oil at the agreed fixed price but for this month
their income will be reduced by the net swap settlement amount. Even though they
have incurred a loss the example illustrates that on a net basis they are selling oil on a
floating price basis.
Table 6.11 summarises the main types of swap and their respective settlement
prices.
Table 6.11 shows certain swaps, which mirror the underlying physical markets with
respect to locations and delivery modes, e.g. Barges FOB Rotterdam. In the case of such
a price assessment, the first criteria refers to the type of delivery vessel (Barges), the
middle term (FOB – Free on Board) relates to how the delivery will be priced, whilst
the third criteria refers to the delivery location (Rotterdam).
Table 6.11 shows the main type of fixed for floating swap by product basis; there
are several alternative ways in which the cash flows can be calculated:
▪ Fixed price vs. quoted futures price: these ‘futures swaps’ are usually traded against
the ‘front’ month and might be applicable for those who have priced their physical
deliveries against either basis but wish to transform it to the other. Due to the expiry
dates of the futures, which are typically during a month rather than at the end, the
price of a swap will consist of futures prices from two different months.
▪ Index price vs. index price: this might be executed as a deal based on two different
indices. A possible example might be Dated Brent against Urals. A swap done on
these indices would be structured in the traditional fixed vs. floating format but
the fixed price will take its value from the differential between the two indices. The
floating payment will be based on the actual differential that is observed over the
agreed payment period for the swap.
218
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 6.11 Frequently traded crude oil and refined product swaps.
Underlying Commodity
Settlement price
Brent
Average of daily settlement price of prompt ICE futures
contract.
Average of daily published S&P Global Platts assessment.
Average of daily settlement price of prompt CME futures
contract.
Average of daily published S&P Global Platts assessment.
Average of daily published Asian Petroleum Pricing
Index assessment.
Average of daily published Argus assessment.
Dated Brent
WTI
Dubai
Tapis
Gasoline 10 ppm 95 Ron Barges
FOB Rotterdam
CME RBOB Gasoline
ICE gas oil
CME Ultra Low Sulfur Diesel
Gas oil 0.1% Sulfur Barges FOB
Rotterdam
Gas oil 0.25% Sulfur Cargoes
FOB Singapore
Fuel oil 3.5% Sulfur Barges
FOB Rotterdam
Fuel oil 2% Sulfur Cargoes FOB
Singapore
Average of daily settlement price of prompt CME futures
contract.
Average of daily settlement price of prompt ICE futures
contract.
Average of daily settlement price of prompt CME futures
contract.
Average of daily published S&P Global Platts assessment.
Average of daily published S&P Global Platts assessment.
Average of daily published S&P Global Platts assessment.
Average of daily published S&P Global Platts assessment.
Note: For the gasoline contract 10 ppm indicates there are 10 parts of lead per million.
As to when someone would use this type of swap, a possible scenario would
be where a refiner buys a cargo of Urals and decides to hedge it with Brent futures
given the absence of a futures market in his particular cargo. To hedge the basis
risk that exists between the two cargos, the refiner enters a swap to lock in the
differential between Urals and Dated Brent. Although it may seem easier for the
hedger to do a fixed/floating based purely on the price of Urals, this would not be
executed due to poor liquidity. The hedger is then exposed to a change in the Dated
Brent–Brent futures differential.
▪ Index price vs. futures price: this type of swap would be quoted as an index price
against a futures price, e.g. Dated Brent vs. ICE Brent futures.
▪ Crude oil vs. products: equally there may be swaps based on different refined products. For example, the swap could involve the exchange of cash flows based on the
price of crude oil against one of the refined products such as gasoline. Equally the
contract might be structured with cash flows based on the price of fuel oil, against
that of electricity.
219
Crude Oil
Hedging swaps with futures
Dealers who enter into long-dated swaps in an illiquid product may be forced to hedge
the exposure using crude oil futures. This will give them what is sometimes referred to
as a ‘crack position’ (i.e. an exposure in a refined product against crude oil). For example,
suppose that a bank executes a jet fuel swap with an airline such as the one illustrated
in Figure 3.1. In simple terms, the bank will lose money on the swap, if the price of jet
fuel were to increase above the fixed price. A possible hedge would be to buy crude oil
futures. The logic being that if jet fuel prices rise, then (hopefully) crude oil prices will
rise to generate a profit on the futures position to offset the loss on the swap.
This, however, raises a few questions:
How many futures does the trader need to execute? One possible answer to this question would be for the trader to perform some form of regression analysis. This technique describes the relationship between a dependent and independent variable. In this
instance the dependent variable would be the price of jet fuel whose value is predicted
based on movements in the independent variable, which would be the crude oil future.
The first step would be to create a scatter graph of the data and to determine a ‘line
of best fit’ whose mathematical form would be:
Δy = 𝛼 + βΔx + ε
Where:
Δy = the change in jet fuel prices (the dependent variable)
𝛼 = a constant; the value of Y even if X has zero value
β = the regression co-efficient. This represents the slope of the line of best fit.
Δx = the change in crude oil futures prices (the independent variable)
ε = the error term.This reflects the fact that other factors may influence
the value of Y and are not captured by the specification of the equation.
Suppose that the regression equation derived from the data is:
Y = 0.001244 + 0.50x
If, for ease of illustration, the value of the constant is ignored then the regression
result shows that a one-unit change in X (i.e. crude oil prices) is associated with a
0.50-unit change in Y (i.e. jet fuel prices) over the sample period. Admittedly, the values used in the equation were used to illustrate the principles. If jet fuel prices move by
USD 1.00 this would be associated with a USD 2.00 move in crude oil prices. Since the
crude price is more volatile than jet fuel price the trader would need a smaller futures
position to hedge the jet fuel exposure.
220
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
There are problems associated with this technique.
What is the reliability of the regression result? The value of the regression co-efficient
will also be a function of the sample period. In addition, there is no guarantee that the
historical relationship between the two variables will extend into the future.
Which maturities should be traded? Broadly speaking, a trader could execute either
a ‘stack’ or ‘strip’ hedge. A stack hedge would require the trader to buy all the required
futures in a single maturity, which may well be the nearest dated contract. Since the
maturity of the swap would be greater than the maturity of the hedge, this position
would need to be ‘rolled’ as the future approaches maturity. The rolling procedure
requires the trader to close out the existing position and then reopen a new position
in the next maturity. In this case, since the trader is long the future, they would need
to sell futures to terminate the exposure and then buy futures in the next maturity to
maintain the hedge. A ‘strip’ hedge would consist of a series of futures with sequential
maturities spread out over the life of the underlying swap.
Which strategy is more effective is a debatable issue. Although a strip hedge mirrors the maturity of the underlying swap it may be that there is no liquidity in certain
maturities, i.e. although the trader may wish to buy a future, there may not be a willing
seller.
How effective will the hedge be? Hedging a jet fuel exposure with crude oil is not going
to be a perfect hedge. It does assume that the two energy products will be positively
correlated by virtue that there is a chemical relationship between the two. Although
there is a strong reason to suggest this may be the case, there is no guarantee that this
price relationship will hold in all circumstances.
How will the hedge impact the price of the swap? There is an old market adage that
says the price of a derivative is a function of the underlying hedge. So if the trader offers a
jet fuel hedge to a customer, who is hedged somewhat imperfectly with crude oil futures,
the trader may decide to make an adjustment to the fixed price receivable on the swap
to protect them from an adverse move in the value of the hedge.
Two-asset barrier options
Two-asset barrier options were introduced in the gold chapter but to show their
flexibility, they are illustrated here within a different context. In this section, the
scenario is based on a non-US crude oil producer who will receive USD revenues but
may wish to convert them to another currency such as EUR. The conventional way of
doing this would be using a standard FX option. A stylised term sheet for a EURUSD
FX option would have the following terms:
Notional:
EUR 50 m
Reference currency pair:
Maturity:
Current spot rate:
Strike:
Premium:
EURUSD
One month
EUR 1.00 = USD 1.1500
EUR 1.00 = USD 1.1524 (ATM forward)
USD 0.0106 per EUR of notional (50 m × USD
0.0106 = USD 530,000)
221
Crude Oil
For readers unfamiliar with FX option market conventions there are a few points
to note:
▪ It is not sufficient to refer to an FX option as a EURUSD call in the same way one
might with a crude oil call. Conventions vary between FX markets so the term call
could refer to the base currency (EUR in this case) while in other markets it may
refer to the quoted or counter currency. Since the purchase of one currency requires
the sale of another then to avoid confusion it is better to say that in this instance
the producer has purchased a EUR call-USD put.
▪ Like all options the premiums are quoted in the same units of the underlying. A
EURUSD exchange rate expresses the number of USD to buy or sell a unit of EUR.
Therefore, the premium is the number of USD per unit of EUR. In this case it is just
over one cent for every EUR. Since there are 50 m units of the base currency then
the premium is scaled appropriately to give a value of USD 530,000.
▪ If the client wishes the premium to be expressed in the base currency, then this
is converted using the spot rate. In this case it would generate a premium of EUR
460,870.
▪ Some clients may also like the premium to be expressed as a percentage of the
notional amount so in EUR terms it would be 0.92% (EUR 460,870/EUR 50 m).
One of the ways to reduce the premium cost of this FX hedge would be to structure
the option to knock out when commodity prices are high. With high crude oil prices
producers are likely to be cash rich and maybe willing to sacrifice some degree of FX
protection. Based on the previous figures a two-asset barrier which knocks out with
high oil prices may look as follows:
Notional:
EUR 50 m
Asset 1:
Maturity:
Current FX spot:
Strike:
Asset 2:
Asset 2 barrier:
Current crude price:
Payoff:
EURUSD
One month
EUR 1.00 = USD 1.1500
EUR 1.00 = USD 1.1524 (ATM forward)
Brent Crude Oil
USD 65.00/bbl.
USD 60.00/bbl.
If the FX spot rate at maturity is greater than EUR 1.00 =
USD 1.1524 AND the Brent crude price is less than USD
65.00/bbl., then option buyer will receive EUR 50 m in
exchange for USD 57,620,000. Otherwise no exchange
will take place.
From the producer’s perspective since they are receiving USD, they are at risk that
this currency will depreciate relative to the EUR. A USD depreciation (weakening)
means that the exchange rate will move in such a way that they will have to give up
more USD per EUR.
222
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The expected correlation between the two assets is a crucial component of this type
of option. The following premiums would apply depending on the value of the correlation input:
−1
0
+1
USD 0.0106
USD 0.0101
USD 0.0080
There are some interesting features to highlight from this structure:
▪ The correlation input measures the strength of the relationship between the crude
oil and the underlying currency, which is EUR not USD.
▪ A negative correlation of −1 means that the option trades at the same premium as
the vanilla FX option. This is the correlation between the EUR (not the USD) and
the price of crude oil. So, if the EUR strengthens (the USD weakens) this will always
be associated with a fall in the price of crude oil. Therefore, the option will never
knock out and so will be indistinguishable from the vanilla structure.
▪ A positive correlation of +1 means that if the EUR weakens (USD strengthens)
this will be associated with a decrease in the price of crude oil and so the option
will not be knocked out. Although the option will not get knocked out, it will not
be exercised, as the FX component will be out-of-the-money. In the spot market
the FX rate will be less than the strike meaning the producer would be better off
walking away from the option because in a spot deal they would give up fewer USD
per EUR.
▪ The monitoring of whether the price of crude oil breaches the barrier is mostly done
on a continuous basis. If, however, it were a ‘European’ barrier when it was only
monitored at maturity then using the zero-correlation premium as an example, the
cost of the option would increase. Since the barrier is only monitored once during its
life, the chances of crude oil being above the barrier on the expiry date is relatively
lower than had it been monitored on an on-going basis. This reduces the likelihood
that the option would be knocked out, making a payout more likely and therefore
increasing its cost.
▪ The barrier on this structure is relatively close to the current price of crude oil. If
the barrier were to be increased, the cost of the option would increase as the chance
of it being knocked out reduces. One of the reasons for this is that when pricing the
option an implied volatility of 20% p.a. was used. A trader’s rule of thumb would be
to convert this value to a monthly equivalent and apply it to the current spot rate to
see the range of values that are expected to trade over the life of the option. Since
the option has a one-month expiry it is tempting to divide the annual volatility by
12, but the correct approach is to divide by the square root of this value. Since the
square root of 12 is approximately 3.46 this would return a monthly equivalent value
of 5.78% (20/3.46). Applying this figure to the current crude oil price of USD 60.00
223
Crude Oil
suggests a range of values of +/− USD 3.47 (USD 56.53 to USD 63.47). So, the closer
the barrier is to the spot price the greater the chance the option will knock out and
so as a result the premium will be lower. Using the same volatility for a one-month
period with a higher barrier of, say, USD 70.00 means that the option is very unlikely
to be knocked out and so will trade closer to the value of a vanilla FX option.
6.9.2
Refiner hedges
The demand for crude oil arises mainly from refineries. In theory, the refiner could
choose to hedge the cost of their crude oil or their income from selling refined products
or both.
Hedging the cost of crude oil
Suppose a EUR-denominated refiner is looking at different swap structures to manage
the cost of their crude oil purchases. One possible solution is an FX-linked swap that
has embedded optionality. A term sheet may look as follows:
Underlying asset:
Current price:
FX pair:
Maturity:
Cash flow frequency:
Fixed price:
Fixed price payer:
Floating price:
Floating price payer:
Reset condition:
FX barrier:
Reset fixed price:
Crude oil
USD 70.00
EURUSD
Two years
Monthly
USD 60.00, subject to possible reset
Refinery
Monthly average of prompt futures price
Bank
If on any date during the life of the swap, the EURUSD
exchange rate is above the FX barrier then the fixed
price on the swap is reset from that point until maturity.
EUR 1 = USD 1.2500 (current spot EUR 1 = USD 1.1500)
USD 85.00
If we assume that the refinery buys its physical crude oil based on an ‘average of
month’ basis that references the futures price, then the floating swap cash flow received
under the swap could be used to finance this payment. As a result, the fixed rate on the
swap would represent their net cost of buying the crude oil. From an FX perspective,
a weakening of the USD (strengthening of the EUR) would generally be beneficial, as
it would mean they would give up fewer EUR to acquire the USD that they require to
buy the physical crude. However, note that in this case, if the FX rate does move in a
favourable manner, the fixed price will reset to a level that is considerably higher than
the current price of crude oil. The refiner’s cost of buying their physical crude oil will
therefore increase.
224
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ The fact that the refiner is able to initially obtain a favourable fixed rate on the swap
that is USD 10.00 below the current price suggests that they have sold some form
of optionality and rather than receive an explicit premium, their ‘payment’ is in the
form of a subsidised fixed rate.
▪ Note that if the rate is reset, this can happen at any time, which suggests a barrier
that is monitored on a continuous basis (a so-called ‘American’ barrier).
▪ The activation increases the fixed amount payable for the remaining life of the swap
to a single higher rate. This suggests that the embedded optionality has a digital type
of payout.
▪ Since the activation of the digital barrier is dependent on an FX rate, it would be
classified as a hybrid structure that will require some form of correlation input for
valuation purposes.
Hedging the crack spread using exchange traded futures
Refiners will be buyers of crude oil and sellers of refined products. Given the nature
of their different inputs and outputs they are exposed to a change in the differential
between the two prices. This is sometimes referred to as the crack spread, which derives
it name from the refining process where the crude oil is ‘cracked’ into a variety of different outputs. The crack spread measures the income earned from the sale of refined
products to the cost of crude oil.
Figure 6.5 shows generic refinery margins in different geographical locations. They
are based on a single crude oil appropriate for that region and optimised refined product
yields based on a generic refinery configuration, also appropriate for each region. The
margins are expressed as a semi-variable basis in that all variable costs (e.g. energy costs
incurred in the production of the products) are included. A positive margin is one where
the income received from the sale of the refined products is greater than the cost of
buying the crude oil. Refinery margins in the USA tend to be higher as they are more
constrained in terms of capacity than Europe or Asia.
The refiner can hedge against a fall in this margin by fixing the spread using a
future, OTC forward or an option. One of the most popular strategies in this area is the
3:2:1 futures crack spread. This is a single transaction that is composed of three crude
oil futures, two gasoline futures, and one heating oil contract. This reflects the fact that
broadly speaking a typical refinery configuration will be such that a barrel of crude will
yield twice as much gasoline than heating oil. As a result, every three barrels of crude
will yield two of gasoline and one of heating oil. Although refineries will also produce
fuel oil and naphtha, there is no futures market for these products and so the exposures
cannot be easily hedged.
To hedge a decline in the margin, the refiner would sell the crack spread, which
would involve a purchase of the requisite number of crude oil contracts and the sale of
refined product contracts (gasoline and heating oil). Note that the action of selling or
buying the spread refers to the nature of the refined product transaction. The spread will
usually have a time dimension to reflect the speed with which the crude oil is refined.
Hence the refiner may sell longer-dated refined product futures than the purchased
crude oil future.
225
Crude Oil
Although the futures hedge will fix the refining spread, it is by no means a perfect
hedge, as the refinery runs two secondary basis risks. The first relates to differences in
the delivery location between the physical transaction and derivative hedge. This will
be reflected in the price of the different contracts and there is no guarantee that they
will move in tandem. For example, assume that a US refiner decides to buy crude oil for
delivery on the US Gulf Coast (USGC) with a similar quality as that of the WTI future
(e.g. Bonny Light). He is likely to sell his gasoline output also for delivery at the same
location (USGC). If futures are used to hedge the refining margin, the crude oil contract
will be priced for delivery at Cushing while the gasoline contract is based on delivery at
New York Harbor (NYH). This means they have exposure to the Cushing/USGC price
differential on the crude oil exposure and NYH/USGC price differential on the gasoline
contract. The second source of basis risk lies in the quality of the underlying physical
products and that specified in the futures contract.
Hedging the crack spread using OTC swaps
It is assumed that a refiner is proposing to hedge its production revenues for a future
fourth quarter. In the physical market they agree to purchase 480,000 barrels of Brent
crude oil according to a uniform pricing schedule based on the daily S&P Global Platts
Dated Brent assessment in October, November, and December. In addition to this, the
refiner agrees to sell 480,000 barrels of its refined products according to several other
published assessments over an identical time frame. The details of the refiner’s production profile are given in Table 6.12.
For the sake of simplicity, it has been assumed that there are no waste products in
this refining process although in practice this would not be the case. We will also assume
the refiner is unwilling to execute a futures-based transaction, as it does not want to be
exposed to the basis risks. Furthermore, the refiner wishes to hedge its entire slate of
products rather than those where a traded market exists.
It is possible to execute an OTC swap on the margin between the weighted basket of
products in Table 6.12 and Dated Brent. Liquid swap markets exist for each of the production constituents, which means that the price of this ‘margin swap’ can be derived
as follows:
Refinery Margin = (0.03 ∗ Naphtha + 0.41 ∗ Gasoline + 0.35 ∗ gas oil
+ 0.21 ∗ Fuel oil) − Dated Brent
TABLE 6.12 Hypothetical refiner’s production profile.
Product
% of total
output
Naphtha
Gasoline
Gas oil
Fuel oil
3
41
35
21
Pricing quote
S&P Platts Cargoes CIF North West Europe
Argus 10 ppm 95 Ron Barges FOB Rotterdam
S&P Platts 0.1% Sulfur Barges FOB Rotterdam
S&P Platts 1% Sulfur Barges FOB Rotterdam
226
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The fixed price on the swap is in essence a weighted average of the component swap
prices, less the price of a Dated Brent swap. The weights applied to the refined products
represent the refiner’s relative production percentages.
In the European markets, refined product prices are quoted in USD per metric
tonne whilst crude oil is quoted in USD per barrel. Refinery margins are quoted in USD
per barrel, which means that the refined product swap prices must be expressed on
the same basis. In the derivatives markets it is standard practice to use the conversion
factors detailed in table 6.13.
The refiner fears that margins will fall by the time fourth quarter arrives and so
enters a margin swap based on a notional of 160,000 barrels per month for the quarter
in question. We will assume that the fixed rate has been set at USD 7.24. The refiner
sells the swap receiving a fixed price of USD 7.24, while paying a floating rate.
Five good business days after each month in the fourth quarter, the refiner pays
or receives a cash amount depending on the difference between the fixed price of
the swap and the average monthly refining margin implied by the various published
prices. Accordingly, whatever it gains or loses in the physical market is offset under
the swap. The result is a hedged position and secured revenues. Table 6.14 shows
how this is achieved and that the net margin achieved is equal to the fixed rate on
the swap.
Hedging the crack spread using options
Basket options were introduced in the chapter on base metals. As a brief recap, a basket
option is a single option that references the value of several underlying assets. One of
the features of the basket option is that they offer a premium saving relative to buying
individual options on each of the underlying assets. This is since the valuation of these
products is a function of the correlation between assets.
The refinery is exposed to a fall in their margins and so two simple option strategies
would be to either buy a put or sell a call on this spread. If they were to buy a put
option, then they would receive a payment if the spread declined but would have to pay
a premium. The sale of a call on the spread would mean that if the spread declined, the
option would not be exercised and the refinery would retain the premium, which could
TABLE 6.13 Energy conversion factors.
Product
USD per bbl. to
USD per metric tonne
Naphtha
Gasoline
Jet
Gas oil
Fuel oil
8.90
8.33
7.88
7.45
6.35
Note: To convert a quantity expressed in USD per barrel to USD per metric
tonne one would multiply by the appropriate conversion factor.
227
Crude Oil
TABLE 6.14 Refinery margin hedge using OTC swaps.
Date
Physical
Swap
15 Sep
Agrees to sell 160,000 barrels per month in
Q4 at average of daily, published
assessments. Delivery/settlement on fifth
business day after month end.
Average margin according to October
published prices = USD 6.82/bbl.
Cash received = (160,000*6.82) = USD
1,091,200
Average margin according to November
published prices = USD 6.03/bbl.
Cash received = (160,000*6.03) = USD
964,800
Average margin according to December
published prices = 8.11 USD/bbl.
Cash received = (160,000*8.11) = USD
1,297,600
Sells (i.e. receives fixed)
160,000 bbl. per month Q4
refinery margin Swap = USD
USD 7.24/bbl.
Swap settlement:
= 160,000*(7.24 − 6.82) = USD
67,200
5 Nov
5 Dec
5 Jan
Swap settlement:
= 160,000*(7.24 − 6.03) = USD
193,600
Swap settlement:
= 160,000*(7.24 − 8.11)
= USD −139,200
Effective sale price = (1,091,200 + 964,800 + 1,297,600 + 67,200 + 193,600 −
139,200) / 480,000 barrels = USD 7.24/bbl.
be used to offset any decline in revenues. If the spread were to increase, then they would
be required to make a payment on the option although their underlying margins would
improve.
A refining margin option is a spread option between a basket of energy products
and crude oil. Instead of trading a series of spread options between each individual
refined product and crude oil, the refiner can trade a single refining margin option,
which would include all its oil products. Given the fact that every refinery will have a
different configuration then these products will usually be bespoke.
This section focuses on the use of calendar spread options for managing storage
capacity. Owning storage gives the owner the choice between selling the commodity
now or in the future. Owning storage capacity can be thought of as a form of ‘real’
optionality.
Suppose the market is in contango with forward prices rising with respect to maturity. There may be an incentive to store the asset now and sell it later at a higher price.
This would be profitable if the costs associated with storing the asset are less than the
difference between the spot and forward prices.
If the market is in backwardation then short-term prices are higher than long-term
prices, and the owner should logically sell the asset immediately as there is no benefit
to leaving it in storage.
This difference between two maturities on a forward price curve is known as the
calendar spread. The maturities could be spreads for either adjacent or non-adjacent
months. It follows that a calendar spread option is an option on the price differential
228
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
between two delivery dates of the same commodity. So, for example, a spread option
could be referenced to the price difference between the June and December ICE Brent
futures contract.
By convention, the calendar spread is defined as:
Near-dated futures price minus the long-dated futures price
Sometimes reference may be made to buying or selling the spread. In this instance,
buying the spread means buying the near-dated contract and selling the long-dated contract. The opposite is true when describing the sale of the spread.
One implication of defining the spread in such a manner is that in a contango
market, the spread is negative, while in a backwardated market the spread is positive.
Representing this diagrammatically shows that the availability of storage can be thought
of as a put option on the calendar spread.
Increasing contango
10
Increasing backwardation
Storage benefit
8
6
4
2
0
-10
-8
-6
-4
-2
0
2
4
6
8
10
FIGURE 6.21 Payoff from a calendar spread option.
So, calendar spread options can be used to protect clients against adverse changes
in the forward curve.
Monetising storage capacity
The forward curve for gas oil typically displays an element of seasonality. That is, they
display a succession of contango and backwardation periods, with the fuel being more
expensive in winter and cheaper in the summer. Suppose it is currently August and
a company has 100,000 cubic meters of excess capacity for the period between January
and June in the following year. The current futures price for January is USD 903.00/MT,
while the June futures price is USD 889.00 indicating that the market is in backwardation. The company decides to sell a put option on the spread between these futures
prices. A put option on the spread would allow the holder to sell or deliver the underlying which in this case is the spread between two futures prices. From an operational
perspective it is often easier to cash-settle such contracts. Calendar spread options can
229
Crude Oil
TABLE 6.15 Monetising storage capacity using options.
January price
June price
Spread (Jan – Jun)
Option strike
Option payout
Gas oil purchased?
Profit and loss on
purchase and sale of
gas oil
Storage cost incurred
Option premium
Total return
Scenario #1
Scenario #2
Scenario #3
Scenario #4
USD 860.00
USD 890.00
− USD 30.00
− USD 10.00
− USD 20.00
Yes
USD 30.00
USD 860.00
USD 868.00
− USD 8.00
− USD 10.00
USD 0.00
Yes
USD 8.00
USD 860.00
USD 866.00
− USD 6.00
− USD 10.00
USD 0.00
No
USD 0.00
USD 860.00
USD 850.00
USD 10.00
− USD 10.00
USD 0.00
No
USD 0.00
− USD 7.00
USD 0.50
USD 3.50
− USD 7.00
USD 0.50
USD 1.50
USD 0.00
USD 0.50
USD 0.50
USD 0.00
USD 0.50
USD 0.50
be structured to be European or Asian in style and generally the option is set to expire
prior to the maturity of the first futures contract9 .
The ‘ATM spread’ would be USD 14.00 (USD 903.00 − USD 889.00). The company
decides to set a strike of −USD 10.00, which means that they will only have to pay out
on the option if the spread declines beyond −USD 10.00, i.e. if the market moves into
contango.
Table 6.15 shows the profit or loss to the company at maturity under a variety of
different scenarios.
The following points are relevant to the table.
▪ The expiry payoff on the option is MAX (Strike − (January price − June price), 0)
▪ If the client takes delivery of the gas oil for the six-month period it is assumed that
the total cost of storage is USD 7.00/MT.
▪ It is also possible that the client may still decide to buy the gas oil for January and
sell the gas oil for June even if the option is not exercised against them. They would
do this if the difference between the buy and sell price is less than the cost of storage.
▪ For ease of illustration the example has ignored any transportation and financing
costs.
▪ The total return to the client is: Payout on the option + storage cost + profit and
loss from the purchase and sale of gas oil + the option premium.
The fact that the strike is negative makes the payoff on this type of option somewhat
counterintuitive.
9
At the time of writing, the ICE gas oil contract matures two days before the 14th calendar day of
the delivery month, with physical settlement taking place between the 16th and the last calendar
day of the delivery month.
230
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Scenario #1 – here the market moves into contango, which pushes the spread of
the two futures prices beyond the strike; recall that a put option will pay off as
the underlying decreases. Since the storage owner is short the option, they will be
required to pay out USD 20.00 to cash-settle the obligation. However, since the market is in contango then as an unrelated separate transaction the storage owner could
now ‘buy the futures spread’ by buying the January contract and selling the June
contract. If they were to hold both futures to maturity they would take delivery of
the gas oil in January for USD 860.00, pay USD 7.00 for the cost of storing it until
June, when they could deliver it to fulfill their short June futures position resulting
in an income of USD 890.00. There overall return would be:
Payout on the option + profit or loss on the futures position − cost of storage
+ option premium
− USD 20.00 + USD 30.00 − USD 7.00 + USD 0.52 = USD 3.50
▪ Scenario #2 – again the market is in contango by USD 8.00. The option expires out of
the money but since the size of the contango is greater than their storage cost there
is still sufficient reason to buy the futures spread as per the first scenario. Using the
same equation as in scenario #1, their overall return would be USD 1.50.
▪ Scenario #3 – the market is in contango, but the option expires out of the money.
However, the amount of contango does not cover the cost of storage and so the physical commodity is not bought and sold. However, since the option is not exercised
the storage owner retains the premium.
▪ Scenario #4 – this is where the market remains in backwardation, the option is not
exercised, there is no incentive to buy and store the physical commodity, and so
similar to the third scenario, the owner earns just the option premium.
In the third and fourth scenarios where the storage has not been used, the owner
has been able to generate some income from the sale of the option, which has not been
exercised.
6.9.3
Refined product hedges
Consumers of crude oil are largely refineries as most end users are more interested in
the final refined product. As a recap, Table 6.16 shows the main refined products, their
uses and applicable industry sectors:
Consider an example of an airline company based in Europe offering flights to a
variety of different global locations. At the start of the year the airline asks several of
the companies to tender for the delivery of jet fuel to different airports from which they
operate (e.g. London Heathrow). The airline has an ongoing need for jet fuel and so
is seeking a fixed number of barrels per month to be delivered. The tender price will
be based on a floating formula and will settle against prices published by S&P Platts.
A normal method of pricing the jet fuel will be based on an ‘average of month’ basis. This
means that at the end of each month, the price paid will be an arithmetic average of the
jet fuel price quoted daily in S&P Platts. However, there are several different quoted jet
231
Crude Oil
TABLE 6.16 Refined products and their main applications.
Refined product
Major uses
Industry sectors
Jet fuel
Road diesel or
heating oil
Naphtha
Gasoline/petrol
Heavy fuel oil
Air travel, military
Road haulage, farming,
space heating
Petrochemical feedstock
Road transport
Seaborne transport, power
generation, asphalt
Logistics
Airlines
Couriers
Utilities
Railways
Shipping
Chemicals
Heavy industry
fuel prices depending on the delivery location. These may include the US Gulf Coast,
North West Europe (NWE), the Mediterranean, and Singapore. The tender will require
the selling company to include the cost of delivery to the airport. The additional cost of
transportation to the airport will be expressed as a premium to the conventional delivery
location. So, the quote may come back from the seller as ‘Average of month based on
a reference price of S&P Platts CIF NWE plus a premium of USD X per barrel to cover
transport costs’.
Using swaps to transform a floating exposure into fixed
If during the year the airline is concerned about a rise in the price of jet fuel, it could
enter a commodity swap to cover part of the exposure. In April, the company executes
the following cash settled jet fuel swap with an investment bank.
Commodity:
Payer of fixed:
Payer of floating index:
Period of swap:
Fixed price.
Notional amount.
Settlement.
Price source.
Floating reference price.
Reference price calculation.
Jet Cargoes CIF NWE9 (Cost, Insurance Freight, North
West Europe)
Airline
Investment bank
Three months commencing 1 October
USD 650.00/tonne
10,000 tonnes per month
Monthly
S&P Platts European
Spot price for Jet Cargoes CIF NWE, published daily
Arithmetic average of the floating reference price over
each month
At the end of each month the average of all of the daily spot prices is calculated and
compared to the fixed price.
10
Another popular index is Jet Barges FOB Rotterdam.
232
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ If the average price is greater than the agreed fixed price of USD 650.00/tonne, the
airline will receive a cash payment based on this amount multiplied by the agreed
number of tonnes.
▪ If the average price is the same as the fixed price, then no settlement takes place.
▪ If the price of jet fuel is on average lower than the fixed rate then the airline will
make a payment for the difference to the bank based on the differential.
Settlement under the swap is normally five good business days after the last day
of the traded month. The payment or receipt under the swap combined with the purchase of the physical commodity will result in a fixed price of USD 650.00/tonne. It is
important to note that in this example, the terms of the swap’s floating reference price,
matched exactly the terms of the airline’s physical purchases.
Hedging jet fuel exposure using gas oil futures
Owing to chemical similarities it is market practice to price physical jet fuel contract
as a differential to gas oil. Jet fuel normally trades at a premium to gas oil as it must
meet higher quality specifications. However, this relationship can break down in certain
cases where supply and demand fundamentals are such that gas oil is more sought after
than jet fuel.
The pricing relationship between the products can be expressed as:
jet fuel = gas oil + jet differential
If gas oil is trading at USD 72.00/bbl., it can be converted to a tonnage equivalent by
using an industry convention value of USD 7.45, which returns a figure of USD 536.40.
If the quoted price of jet fuel is say USD 586.40 per tonne, the ‘jet diff’ is USD 50.00. Gas
oil normally trades at a premium to crude oil and this differential is sometimes referred
to as the ‘Gas Oil’ crack.
It is possible to hedge a jet fuel exposure using gas oil futures due to the existence of
a liquid futures market. Traditionally, there have been very high correlations (i.e. > 90%)
between:
▪ Brent crude oil and gas oil
▪ Brent crude oil and jet fuel
▪ Gas oil and jet fuel
In the gas oil-jet relationship, gas oil has traditionally been the main price driver
and although the relationship can be very volatile, if one were to hedge in gas oil, it is
likely that you would capture most of any jet price movement.
A possible way of pricing a physical delivery of jet fuel could be:
Physical jet fuel = Prompt gas oil futures price at delivery + fixed Jet diff
+ transport costs to delivery location
Crude Oil
233
There could be several ways in which an airline could hedge this exposure:
▪ Transfer the risk to their suppliers by agreeing a fixed price supply contract.
▪ Transfer the risk to customers by including some provision such as a ‘fuel levy’
clause that would require a supplemental payment in the event of an increase in
prices.
▪ Since the physical purchase references a futures price, the airline can buy the requisite number of gas oil futures and with the other two elements of the physical
contract fixed, they will be able to lock in a known future cost.
Another strategy sometimes used by airlines involves a combination of different
derivatives depending on the time horizon.
▪ Long-dated crude oil trades of a three- to four-year maturity to manage the overall
macro picture.
▪ As these contracts shorten in maturity, they may then roll into gas oil futures with
maturities of about two years.
▪ With one year to go they may seek out jet fuel solutions (e.g. swaps) and then with
six months remaining look to hedge deliveries to specific locations.
Hedging jet fuel exposures using crude oil futures
In the previous section, it was argued that airlines might decide to hedge their jet fuel
exposures using crude oil futures if there is no liquid instrument that directly matches
their exposure. Banks that offer risk management solutions to clients may be faced with
a similar problem. For example, suppose that a bank enters a fixed-floating jet fuel swap
with an airline, where they receive fixed and pay floating. Since the bank is now faced
with an exposure to rising jet fuel prices, they need to hedge this exposure, which could
be done using crude oil futures. An example of this type of swap was illustrated in
Chapter 3 (see Figure 3.1) in relation to the collapse of Japan Airlines.
Regression analysis could be used to determine the appropriate size of the hedge.
Regression analysis describes the relationship between a dependent and independent
variable. In this example, the dependent variable would be jet fuel and the independent variable would be crude oil futures. The form of the regression equation would be:
Change in the price of jet fuel = Alpha + Beta Change in crude oil futures price + e
Where:
Alpha = a constant; the value of Y even if X has zero value.
Beta = the regression coefficient. This is the slope of the line of best fit.
e = error term. This reflects the fact that other factors may influence the
value of the dependent variable and are not captured by the
specification of the equation.
234
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Very often in financial price regressions the value of alpha can be small and is often
ignored. It is the value of beta that is used to calculate the appropriate hedge ratio.
To calculate the equation a few assumptions need to be made:
▪ Which index value of jet fuel should be used?
▪ How far back should the data sample extend?
▪ What is the frequency of the sample? (Daily, weekly, monthly)
▪ At what time of day are the futures prices sampled?
▪ Which maturity of futures should be used?
▪ What conversion factor do we need to convert jet fuel (traded in metric tonnes) to
crude oil (traded in barrels)?
With respect to the final point a common factor used in the energy markets is that
one metric tonne is equivalent to 7.45 barrels. If jet fuel is trading at USD 650.00/MT,
then this is equivalent to USD 87.24/bbl.
Suppose that the results of the regression analysis suggest a beta value of 0.8. This
would tell us that a one-unit change in the price of a crude oil future is associated with
a 0.8-unit price change in jet fuel. The bank decides to hedge 1,000 metric tonnes (MT)
of jet fuel using crude oil futures with a contract size of 1,000 barrels. The correct hedge
ratio would be:
= Beta ∗ (jet fuel amount to be hedged∕contract size of crude oil futures)
= 0.8 ∗ (7, 450 barrels of jet fuel∕1, 000 barrels of crude)
= 6 crude oil futures (rounded to the nearest contract)
If crude oil futures were to increase by USD 1.00 then the futures hedge would
change in value by USD 6,000. Based on the regression analysis, this dollar change in
the value of crude would be associated with a USD 0.80 change in the price of jet fuel
resulting in a change in value of USD 5,960. The hedge is not 100% perfect because an
element of rounding was used to determine the correct number of futures.
Although the calculation may appear to be somewhat straightforward, the trader
is then faced with some tricky issues. If hedging a multi-period swap, which maturity
of futures do they pick? They could trade a strip of futures (i.e. a series of consecutive
futures contracts) or a stack of futures (hedge the entire maturity of the swap by using
one single futures maturity). A strip of futures may closely match the underlying exposure but may not be liquid. The second approach offers the greater liquidity but does
not match the maturity of the swap. Another problem is that the beta value will likely
evolve over time and so any hedge will need to be rebalanced to reflect any change.
Hedging jet fuel exposures using gas oil futures and a basis swap
There is an active OTC market for basis swaps that allow an end user to transform exposure in one product to that of another, i.e. an underlying exposure to gas oil can be
transformed into a jet fuel exposure. These swaps are quoted on a bid-offer basis and
are quoted as a fixed against floating differential, in the same manner as a Brent CFD.
The fixed price is calculated as the difference between gas oil and jet fuel.
235
Crude Oil
In this example, we will assume that the airline buys their jet fuel requirements in
the spot market on a variable price basis (e.g. an average of quoted S&P Platts prices plus
a premium for delivery to a particular location). To hedge the exposure, they decide to
use a combination of gas oil futures and a basis swap. They buy the gas oil future with
the same maturity as the underlying exposure at the equivalent of USD 536.40/tonne.
At the same time, they enter a jet diff swap at a quoted level of +USD 45.00. The swap
is bought in the sense that the agreement is requiring them to pay a USD 45.00 fixed
amount per tonne at expiry in return for receiving the average difference between the
actual gas oil price and jet fuel price over the agreed period of the swap.
The combination of the two deals (buy the gas oil future, buy the jet diff swap)
hedges the two components to which they were exposed under the terms of the physical.
Under this pricing agreement the buyer is exposed to a change in the price of gas oil and
the jet differential since jet fuel is priced from gas oil. The net effect of the transaction
is that the airline locks in the price of their jet fuel purchase at USD 581.40.
Assume that at maturity the following market conditions exist. The gas oil future is
trading at USD 550.00/tonne and the jet differential has narrowed to USD 40.00. From
the relationship presented earlier that links gas oil and jet fuel, we can infer that jet fuel
is now trading at USD 590.00/tonne. The airline will make a USD 13.60 profit on the
close out of the future but will have to make a net payment of USD 5.00 under the swap.
The airline buys jet fuel at USD 590.00/tonne to which they add the USD 5.00 loss on
the swap but subtract the USD 13.60 profit under the future. The net cost is USD 581.40,
equal to the locked in value established at the point of executing the gas oil future and
buying the swap.
Hedging jet fuel using options
Suppose an airline consumes 1,000 MT of jet fuel each month but which varies by
100 MT over or under this amount. The base amount of their demand could be hedged
using a series of call options (a ‘strip’ of options), which could be structured in a variety
of ways highlighted previously in the text (e.g. Asian-style, premium reduction techniques). The variable demand could be hedged using a compound option. A compound
option is an option on an option so in this instance the airline could buy a call on a call.
They would pay a premium to acquire a call option, which if exercised would allow
them to buy a second call option. On the exercise of the second option, the airline
would be obliged to pay a second premium.
The terms of a plain vanilla European call option would be:
Underlying price:
Strike:
Time to maturity:
Implied volatility:
Premium:
USD 600.00
USD 600.00
One month
30%
USD 21.00 (rounded)
236
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
A compound option would allow the airline to buy the above option based on these
terms. So, the call on a call would have the strike rate equal to the premium on the
second, plain vanilla call. The terms of the compound option would be:
Underlying price:
Strike on underlying option:
Strike on compound option:
Compound option maturity:
Vanilla option maturity:
Implied volatility:
Premium:
USD 600.00
USD 600.00
USD 21.00
One month
One month
30%
USD 10.00 (rounded)
In this example, the compound option has a one-month maturity which if exercised
grants the holder a one-month call option. For the purposes of this example, both dates
were chosen somewhat arbitrarily and can be negotiated between the counterparties.
When the compound option matures, the airline would need to make a judgment
as to their likely consumption in one month’s time. If they anticipate consuming the
extra jet fuel, they may decide to exercise the compound option and enter the vanilla
call option. This would only be worthwhile if the additional premium now represents a
saving over the current vanilla equivalent. So, if implied volatility had increased over the
month (all other things unchanged) a one-month plain vanilla call option would now
cost more, say, USD 24.00. In this scenario the compound option would be exercised, as
it is ITM. However, note that the total premium incurred by using the compound option
(USD 31.00) is greater than the now more expensive call option (USD 24.00), which in
hindsight would make this approach a relatively expensive one.
CHAPTER
7
Natural Gas
7.1
FORMATION OF NATURAL GAS
Natural gas is a fossil fuel, the main constituent of which is methane (CH4 ). Oil and natural gas, which are frequently found together in the same deposits, are formed from the
decay of vegetation and animals. Over time, geological processes turned these remains
into reservoirs of hydrocarbons trapped by overlying impermeable rock strata.
Natural gas that is discovered with crude oil is often referred to as ‘associated gas’
but is classified as ‘non-associated gas’ when found separately. Although the actual composition of natural gas varies between reservoirs, a distinction can be made between
‘wet’ and ‘dry’ gas. Wet gas has a high proportion of other gaseous substances such as
pentane, ethane, propane, and butane referred to collectively as natural gas liquids or
NGLs. Dry gas is natural gas without these associated substances. After natural gas has
been extracted from the ground the NGLs are removed and can be sold separately. For
example, ethane is a key input in the production of plastics. The processing of natural
gas also removes any water and hydrogen sulfide and adds a smell for safety purposes,
as methane in its naturally occurring form is odourless.
A major development in the industry is the significant technological advancements
made with respect to the recovery of natural gas trapped in shale rock formations.
These formations will sometimes have fractures that contain natural gas, but since
the surrounding rocks are not particularly permeable, the extraction process was not
commercially viable for many years. This type of gas is sometimes referred to as an
‘unconventional’ source. Shale gas extraction is not a new concept and the author
recalls a discussion with an experienced energy professional who witnessed shale
operations in the US in the late 1970s/early 1980s. Two relatively recent developments
that have increased the viability of such projects are:
▪ Horizontal drilling – this allows the pipeline to the wellhead to bend, which not only
allows access to difficult-to-reach areas, but also increases the surface area that can
be drilled.
▪ Hydraulic fracturing (‘fracking’) – this involves pumping liquids into the shale rock
formation to create wider fractures in the rock. The liquids may contain a more
permeable solid such as sand, which would then allow the trapped gas to flow to
the wellhead.
237
238
7.2
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
MEASURING NATURAL GAS
The most common measure of the volume of natural gas is a cubic metre, which is
normally expressed as m3 or cm. In the USA, the convention is to use imperial measures
such as cubic feet (ft3 ). These measures are quite small and so are often multiplied by
sizing factors such as a thousand (m), million (mm; 106 ), billion (b; 109 ) and a trillion
(t; 1012 ). These measurements are taken assuming normal temperatures and pressures.
The energy content of natural gas is measured in a variety of ways. The calorific
value of natural gas measures the amount of energy produced when a fuel is burnt.
One calorie measures the amount of energy required to raise one gram of water by one
degree Celsius. A joule is an alternative measure, and one calorie equals four joules.
A kilowatt is equal to 3,600,000 joules and is often expressed as a rate. For example, a
kilowatt-hour (kWh) is equal to one kilowatt of power expended for one hour. Larger
businesses and institutions sometimes use the megawatt hour (MWh), where 1 MWh =
1,000 kWh. The energy outputs of power plants over long periods of time or the energy
consumption of states or nations can be expressed in gigawatt hours (GWh; equal to
1,000 MWh).
In the USA, the most common measure of energy content is the British Thermal
Unit (BTU). It is defined as the amount of heat that is required to raise the temperature
of one pound of water by one degree Fahrenheit. Another common measure in the UK
market is the therm, which is equivalent to 100,000 BTU or about 97 cubic feet. In the
European natural gas markets a variety of different measures can be used. These include
therms, megawatt hours, or gigajoules. One MWh is equal to 34.12 therms and 1 million
BTUs equal about 10 therms.
7.3
THE PHYSICAL SUPPLY CHAIN
One of the effects of natural gas deregulation has been to increase the degree of competition along the physical supply chain. The traditional monopolistic supply chain has
been restructured so that different functions can now be performed by different entities.
While some companies may be vertically integrated performing a number of functions
along the supply chain, others may just operate ‘upstream’ in exploration and production or ‘downstream’ in the trading and supplying of gas.
7.3.1
Production
The first part of the physical supply chain is the exploration and production of natural
gas from either an offshore or onshore location. If the natural gas is produced offshore
it has to be gathered from various rigs and piped ashore. The point at which the natural
gas reaches the shore is sometimes referred to as the beach terminal. A significant number of producers will be oil majors, but there may also be several smaller independent
companies in operation. In some locations there will also be onshore production facilities, but the process will essentially be the same. Before onward transportation this ‘wet’
natural gas will have to be processed to separate the natural gas element (i.e. methane)
from the other naturally occurring natural gas liquids. The resultant ‘dry’ gas is then
delivered into the pipeline system.
Natural Gas
7.3.2
239
Shippers
The advent of deregulation and subsequent re-regulation has allowed third party access
(sometimes called ‘open access’ in the USA) to the National Transmission System (NTS)
and also saw the emergence of a new role within the supply chain; that of the shipper (sometimes referred to as marketers). Shippers are licensed wholesalers who buy
natural gas from the producers at the shore terminals for onward delivery along the
main pipeline to end consumers. The ownership of designated shippers is diverse. They
may be affiliated to natural gas producers, electricity companies, banks, or commodity
traders.
The shipper/marketer will perform a variety of different roles:
▪ Buy natural gas from producers and then find sellers.
▪ Sell natural gas to customers and then source the supply.
▪ Arrange transportation of natural gas along the network pipeline.
▪ Inform the owner of the NTS (e.g. National Grid Gas in the UK) where the natural
gas will enter the pipeline and where it will exit. This is done using a nomination
system. A shipper can nominate gas to be delivered into the system, out of the system, or to simply change the title of gas held within the system.
▪ For natural gas to be shipped along the main high-pressure NTS, the shipper has to
acquire sufficient entry capacity to permit the flow and then nominate the amount
they wish to deliver. Equally to remove natural gas, exit capacity needs to be booked.
Entry and exit capacity can be booked directly with the system operator, via an
auction or by secondary market trading.
▪ Natural gas can enter the NTS from offshore or onshore production facilities or
storage. It exits the NTS into local distribution networks, storage facilities, or is
delivered to large industrial users. As part of their responsibilities the shippers must
ensure that their daily operations balance. That happens when the amount of natural gas injected into the system equals the amount withdrawn. Shippers who fail
to balance their demand and supply incur financial penalties.
7.3.3
Transmission
Once onshore, natural gas is principally moved along a high-pressure pipeline to either
a natural gas retail supplier or directly to a wholesale consumer. By transporting it under
pressure a greater volume of natural gas can be moved. In order to keep the natural gas
moving along the pipelines there will be a number of compressor stations that ensure
that the natural gas remains pressurised. A natural gas retail supplier will arrange for
the natural gas to be delivered to the end user along lower pressure pipes that join the
high-pressure pipeline. These pipelines are usually made from carbon steel and can vary
in size from 2–60 inches in diameter.
One of the consequences of re-regulation was the realisation that new entrants to
the market would not be willing to build competing large-scale national transportation
infrastructure. For example, in the USA the EIA estimate there are about three million miles of natural gas pipelines. Most natural gas users receive their supplies from
an entity that is often referred to generically as a local distribution company (LDC).
240
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Depending on the degree of deregulation in the market, the LDC may operate in a specific geographical area or may be free to offer services throughout a specific marketplace.
In the USA, the delivery points to LDCs are sometimes referred to as ‘citygates’ and are
often used as pricing points for buyers and sellers.
In the UK, National Grid Gas is the company that manages 4,200 miles of the
main high-pressure pipeline; sometimes referred to as the National Transmission System (NTS). It is a private company (also sometimes referred to generically as the Transmission System Operator or TSO), but operates under a licence granted by the UK
regulator. Its role is to deliver natural gas and it can only buy or sell natural gas for purposes of managing the integrity of the pipeline. It is obliged to follow the terms of the
‘Network Code’, which is a framework that dictates how it will operate and the nature
of the relationship it has with the users of the network such as shippers. The Network
Code enshrines a principle that is common to re-regulated markets which is ‘third party
access’. Third party access allows certain companies (shippers) to have access to the
pipeline network in order to facilitate the movement of natural gas on behalf of producers or consumers. There are also a number of independent natural gas transporters
who have built extensions to the NTS to supply certain customers not served by the
main network.
Natural gas leaves the NTS on a continual basis either directly to large industrial
users (such as power stations), storage facilities, or into one of a number of Gas Distribution Networks into which the UK is grouped. Once within the regional network, the
natural gas can be routed towards the final consumer.
One of the key responsibilities of the transmission system operator is to ensure that
the system is balanced. This is done as a three-stage process. Firstly, there is the nomination process, where the shippers are required to inform the TSO as to how much natural
gas will be delivered, over what period and the entry and exit points. The TSO will then
confirm the nomination after they are sure that sufficient capacity exists within the network. The final stage in the process is the scheduling by the TSO of all the confirmed
nominations based on defined priorities.
Although the act of balancing the network will be performed by the TSO, the shippers will be incentivised through monetary penalties to ensure that the amount of natural gas entering the system matches the amount of natural gas exiting. The transmission
system operator has a number of options available to them in order to ensure that the
volume of natural gas within the pipeline is kept within acceptable limits. For example,
they could utilise production contracts to supply natural gas that include a clause that
allows for the contracted amount to be increased above the stated amount. This additional amount is referred to as the ‘swing’. The existence of adequate storage facilities
is also important so that natural gas could be withdrawn at short notice. A third tool
that could be used is interruptible supply contracts with large industrial consumers.
Here the customer is willing to accept that his supply of natural gas may be curtailed
in periods where demand is high. In exchange, the consumer will receive some form of
financial compensation.
7.3.4
Interconnectors
Gas interconnectors connect gas transmission systems from other countries (e.g. Belgium, Netherlands) to the National Transmission System (NTS) in England, Scotland,
Natural Gas
241
and Wales. As production in the North Sea declines, the importance of the interconnectors will increase, as the UK will be forced to import more natural gas.
7.3.5
Storage
Storage plays an important role on the supply side of the market by allowing participants to meet peak demand and may also help ‘smooth’ out demand either during
certain times of the year (e.g. winter) or perhaps during the day (e.g. meeting residential
demand).
Excess natural gas can be stored in large underground caverns, above ground facilities or within the pipes on the transportation network (a process referred to as ‘line
packing’). By raising and lowering the pressure on any pipeline segment, a pipeline
company can use the segment to store gas during periods when there is less demand
at the end of the pipeline. Using line packing in this way allows pipeline operators to
handle short-term fluctuations in demand.
There are different types of storage facilities. A depleted gas field is a very common and cost-effective way of storing gas, as it will already have a significant amount
of existing infrastructure. Underground cavities are created by drilling down into salt
layers, adding seawater to dissolve the salt and then pumping in natural gas to force out
the water. Depleted oil or natural gas reservoirs are also used for this purpose. Natural
gas can also be stored in a liquefied form (LNG), which reduces the amount of space
occupied.
Storage is sometimes classified as either base or peak load in nature. Base load storage is more strategic in nature and although the facilities may be large in size, the rate at
which the gas can be extracted is low. Natural gas reservoirs that have been depleted are
a popular choice for this sort of facility. Peak load facilities are more tactical in nature
and are designed to allow smaller amounts of gas at a faster rate. Salt caverns are commonly used for peak storage needs.
In a re-regulated market, storage facilities are run on a commercial basis with capacity being sold off to third parties. The availability of storage capacity relative to demand
will have an important impact on the price of natural gas. If capacity is scarce relative to
demand, a cold spell of weather could cause natural gas prices to rise steeply. In general
terms, excess natural gas is injected into storage during warmer periods and withdrawn
in colder times.
7.3.6
Supply
Under monopoly-style structure the final consumer would have virtually no choice as
to who provided their natural gas supply. In the natural gas markets that have been
re-regulated, a consumer should be free to choose who provides their natural gas supplies. This freedom of choice also extends to the suppliers who can choose the source
of their natural gas and arrange with a shipper to ensure it is delivered.
7.3.7
Customers
Distinction is usually made between two different types of customer. Retail consumers
would typically include households, shops, and offices whereas very large industrial
242
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
customers would cover power stations that may buy directly from either shippers or
producers. In some cases, the large industrial users may have separate arrangements
to receive their supplies by a specific ‘offtake’ from the national pipeline system.
However, within the industrial capacity there will be larger customers who have an
arrangement whereby their supply can be interrupted or where the agreement to supply
gas is ‘firm’.
7.3.8
Non-physical participants (NPPs)
The re-regulation of physical markets and growth of financial risk management products has seen the emergence of participants who do not have a need for the physical
commodity. Typically, this would include commodity trading houses, hedge funds, and
some banks. There are a variety of roles that NPPs can fulfill:
▪ Traditional lending banks that have lent to upstream producers take on an element
of commodity risk. If they cannot offset that risk, they may seek out an institution
with natural gas trading capability to manage the associated price risk.
▪ Most of the oil producers (e.g. BP, Shell) will have very sophisticated natural gas
trading and risk management operations and will act as trading counterparties to
other NPPs.
▪ Utilities will have large demand and uncertain consumption forecasts and so will
have significant volume and price risk. As a result, they require structured natural
gas risk management contracts that provide them with flexibility.
▪ Major industrial natural gas consumers are chemical companies, building products, glass, steel manufacturers, and fertiliser manufacturers. Where an entity has
a multisite operation, it may also be possible to aggregate smaller loads and manage
the natural gas price risk centrally.
▪ Commodity traders and hedge funds will use the natural gas market to implement
view driven strategies using the suite of physical and derivative products.
7.4
DEREGULATION AND RE-REGULATION
The emergence of natural gas as a tradable commodity has come about because of deregulation in some countries. This, however, does not mean that the various markets are
now unregulated, rather a new set of regulations have subsequently emerged.
Prior to deregulation, gas markets were characterised as monopolies where one
entity would purchase virtually all of the domestic production, transport it to its end
destination, and then sell it to a final consumer. Agreements to buy and sell were primarily based on long-term, fixed-price contracts, often linked to the price of oil. In the
US, prior to deregulation, natural gas producers would sell to pipeline companies who
then transported and sold the commodity either to large industrial users or to local
distribution companies for onward sale to the general public. The prices charged by
producers to the transportation companies and the amounts paid by the local distribution companies were regulated at the federal level. The retail consumer was protected
against monopolistic practices by state regulation.
Natural Gas
7.4.1
243
The US experience
The first country to introduce competition into their domestic natural gas market was
the USA. In the very early days of the industry it was regulated at a local level, but as
intrastate pipelines started to develop there was no effective method of overseeing the
industry. The origins of formal federal regulation were set out in the Natural Gas Act of
19381 , which recognised that while the movement of natural gas along major interstate
pipelines was a natural monopoly, it required an element of regulation. The act concerned itself with the setting of rates for the transportation of natural gas along interstate
pipelines, but did not specify any particular regulation of producer prices. Wellhead
price regulation came about in 1954 as a result of a legal case (Phillips Petroleum Co.
vs. Wisconsin). As a result of this lawsuit the Federal Power Commission (who regulated interstate natural gas sales at the time) imposed price ceilings on the amount that
producers could charge, which remained in place until 1978. There were three effects
of the Phillips decision:
▪ Since the sale price for which natural gas could be sold was regulated, there was
little incentive for the producer to devote extra capital to expand exploration.
▪ The price of intrastate gas was not subject to the same regulation and so created a
bias, which encouraged producing states to sell the commodity locally rather than
nationally. This created natural gas shortages in the mid-1970s in some states.
▪ As a result of the Middle East Oil embargos in the 1970s, natural gas became a
more attractive energy substitute. Combined with the relatively low price ceilings
imposed by the regulator, the demand for natural gas increased substantially.
The next significant act was the Natural Gas Policy Act of 1978, which aimed to
create a single market that did not favour intrastate over interstate sales and would
allow market forces to determine the price of natural gas. The act also saw the Federal
Power Commission replaced by FERC (Federal Energy Regulatory Commission). FERC
is responsible for the regulation of interstate natural gas pipelines and storage facilities.
It is also responsible for approving natural gas and liquefied natural gas (LNG) import
and export facilities. The complete unbundling of natural gas services continued via a
series of orders issued by FERC and federal legislators. In 1985, FERC Order 436 (sometimes referred to as the ‘open access order’) required that natural gas pipelines provide
open access to transportation services, albeit on a voluntary basis. However, it was the
Natural Gas Wellhead Decontrol Act of 1989, which effectively led to the restructuring
of the natural gas industry and encouraged market determined gas prices. This meant
the large-scale consumers of natural gas were able to negotiate prices directly with producers and then contract separately for its transportation. FERC order 636 issued in
1992 made the voluntary unbundling of services outlined in order 436 compulsory. This
order stated that the pipeline companies had to separate their transportation and sales
functions to give the customer greater choice as to the sale, transportation, and storage
of natural gas.
1
Subsequently amended by the 2005 Energy Policy Act.
244
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The overall impact for the USA was that producer prices were no longer subject
to regulation and were driven by the fundamentals of supply and demand. Producers
and consumers could contract directly with the pipeline companies no longer taking
possession of the natural gas. They would now only offer transportation as a service
to another company. It also gave rise to a range of new services such as shipping and
trading.
7.4.2
The UK experience
In the UK, the natural gas market was subject to major deregulation and re-regulation
from the mid-1980s (e.g. the Gas Acts of 1986 and 1995). Prior to this time the market
had been dominated by a single monopoly – British Gas. The company was initially
privatised and by the late 1990s had demerged with a series of separate companies
performing different roles (e.g. BG focusing on extraction, National Grid Gas maintaining the transportation network, and Centrica supplying natural gas to retail
customers under the British Gas brand). The Department for Business, Energy, and
Industrial Strategy (BEIS) is responsible for setting energy and climate change policy
and establishes the framework for achieving different policy goals. The industry’s
upstream activities (e.g. exploration and production) are regulated by the Oil and Gas
Authority (OGA), while the downstream activities (e.g. consumer supply) are currently
regulated by OFGEM (Office of Gas and Electricity Markets).
7.4.3
Continental European deregulation
The European market is mainly regulated by a series of directives2 that should eventually be incorporated into domestic law within each member state. The second Gas
Directive of 2003 (which repealed the first Gas Directive of 1998) had the objective of
creating a single unified market for natural gas. This was complemented by the Gas
Regulation of 2005, which expanded on many of the issues outlined in the 2003 Directive. The third European Gas Directive was agreed in 2007 and entered into force in
2009. The directives require each country within the EU to restructure their industry
in order to allow consumers greater choice as to who will supply their natural gas. The
Gas Directives covered a variety of issues such as:
▪ Unbundling of services provided by integrated companies.
▪ Security of supply.
▪ Infrastructure.
▪ Designation of the transmission system operator to run the high-pressure pipeline
network.
▪ Fair access to transmission networks.
▪ Access to storage.
2
An EU regulation is binding in its entirety across all member states. A directive is a legislative
act that sets out a goal that all EU countries must achieve. However, it is up to the individual
countries to devise their own laws on how to reach these goals.
Natural Gas
245
▪ Distribution of natural gas to customers.
▪ National regulation.
Inevitably any market will evolve according to local conditions and at different
speeds. However, a number of common features have evolved for countries that have
introduced an element of competition:
▪ Prices are determined more by conditions of demand and supply rather than being
set by long-term contracts.
▪ Monopolistic structures have been replaced by the unbundling of functions allowing new companies to enter the market and offer greater competition.
▪ New functions have evolved to meet the demands of the new structures (e.g.
shippers).
▪ A new set of regulations has evolved for different elements of the physical supply
chain. For example in the UK the Network Code outlines the relationship between
the owners of the National Transmission System (NTS) and the gas shipping companies who, in a deregulated market, are now able to administratively arrange for
natural gas to be moved along the pipeline.
▪ New traded markets allow the management of natural gas market price risk (both
over-the-counter and on an exchange traded basis).
▪ Greater movement of natural gas across international borders in both gaseous and
liquefied forms.
7.5
THE DEMAND FOR AND SUPPLY OF NATURAL GAS
7.5.1
Relative importance of natural gas
The annual BP statistical review of world energy provides data in relation to the consumption of energy by fuel type. It defines primary energy sources as:
▪ Crude oil
▪ Natural gas
▪ Coal
▪ Nuclear energy
▪ Hydro electric
▪ Renewables
Figure 6.1 showed the relative importance of natural gas indicating that it comprised 24% of overall energy consumption.
The trade in natural gas is characterised by several different features:
▪ Domestic markets where natural gas is moved by pipeline.
▪ The exporting and importing of natural gas internationally by pipeline.
▪ A global seaborne liquefied natural gas (LNG) market.
246
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
S. & Cent. America
4.0%
Europe
1.7%
North America
7.6%
Africa
7.5%
Middle East
38.0%
Asia Pacific
8.9%
CIS
32.3%
FIGURE 7.1 Proportion of proven natural gas reserves.
Source: BP Statistical Review of World Energy 2020. BP PLC.
7.5.2
Reserves of natural gas
Figure 7.1 shows the global distribution of proved reserves. In percentage terms the
three regions with the largest reserves are:
Middle East
CIS
Asia Pacific
38.0%
32.3%
8.9%
The evolution of each region’s proved reserves over the period 1980–2019 is shown
in Figure 7.2.
The two countries with the largest proven reserves are the Russian Federation with
38 trillion cubic metres (Tcm) and Iran with 32 Tcm.
7.5.3
Production of natural gas
The evolution of natural gas production on a regional basis is shown in Figure 7.3. Overall production of natural gas has been rising steadily since 1970 from 975 billion cubic
metres (Bcm) to 3,989 Bcm in 2019. The highest individual producers of natural gas are
the USA (1,128 Bcm) and the Russian Federation (679 Bcm).
247
Natural Gas
Europe
S & C America
N America
Africa
Asia Pacific
200
180
160
140
120
100
80
60
40
20
0
CIS
Middle East
18
20
16
20
14
20
12
20
10
20
08
20
06
20
04
20
02
20
00
20
98
19
96
19
94
19
92
19
90
19
88
19
86
19
84
19
82
19
80
19
FIGURE 7.2 Evolution of proven natural gas reserves; trillion cubic metres 1980–2019
Source: BP Statistical Review of World Energy 2020. BP PLC.
4,500
3,500
S & C America
Africa
Europe
3,000
Asia Pacific
2,500
Middle East
4,000
2,000
1,500
CIS
1,000
N America
500
0
18
16
20
20
14
12
20
10
20
08
20
06
20
04
20
02
20
00
20
98
20
19
96
94
19
92
19
90
19
88
19
86
19
84
19
82
19
80
19
78
19
76
19
74
19
72
19
70
19
19
FIGURE 7.3 Natural gas production 1970–2019, billion cubic metres.
Source: BP Statistical Review of World Energy 2020. BP PLC.
7.5.4
Shale gas
One of the key developments since the publication of the first edition of this text has
been the rapid technological advancements made in recovering natural gas trapped in
deposits deep in shale rock formations. These shale rocks will often have fractures that
contain natural gas and although fracking has existed for many decades, until recently,
it has not been commercially viable. This unconventional source of natural gas has had
a major impact on the supply side of the market. The two main techniques that have
been used are horizontal drilling and hydraulic fracturing (fracking). As the name suggests, horizontal drilling allows the pipeline to the wellhead to bend, which allows the
producer to access sources that are considered more difficult to reach. Fracking involves
pumping liquids and permeable solids such as sand into the shale formation to widen
the fractures, which will then allow the natural gas to flow to the wellhead.
248
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
109
Middle East
76
CIS
Africa
63
S & Cent America
46
Asia Pacific
26
Europe
14
USA
13
0
20
40
60
80
100
120
FIGURE 7.4 Reserves to production ratio.
Source: BP Statistical Review of World Energy 2020. BP PLC.
7.5.5
Reserve to production ratio
One key concern often expressed in relation to energy is that given that fossil fuels are
finite in nature, how long will it be until the resource runs out? The reserve to production ratio (R/P ratio) divides the level of proven reserves at the end of the year by the
production in that particular year. The result is the number of years that the reserves
would last if production were to continue at that level. At the end of 2019, the global
figure was approximately 50 years.
The regional distribution is shown in Figure 7.4.
7.5.6
Consumption of natural gas
The evolution of natural gas consumption on a regional basis is shown in Figure 7.5. The
USA consumes the largest amount of natural gas (846 Bcm), followed by the Russian
Federation (444 Bcm), and then China (307 Bcm).
Typically, the largest users of natural gas are:
▪ The electricity sector.
▪ The energy sector (excluding electricity) including entities such as oil refineries.
▪ The residential sector, which significantly impacts the seasonal price patterns of
natural gas.
7.5.7
Exporting natural gas
The largest pipeline exporters of gas in 2019 were:
249
Natural Gas
4,500
3,500
Africa
S & C America
Europe
3,000
Middle East
4,000
2,500
CIS
2,000
Asia Pacific
1,500
1,000
N America
500
0
19
20
17
20 5
1
20 3
1
20 1
1
20
09
20 7
0
20
05
20 3
0
20 1
0
20 9
9
19
97
19 5
9
19 3
9
19
91
19 9
8
19 7
8
19 5
8
19 3
8
19 1
8
19 9
7
19 7
7
19 5
7
19
73
19 1
7
19
69
19 7
6
19 5
6
19
FIGURE 7.5 Natural gas consumption 1965–2019.
Source: BP Statistical Review of World Energy 2020. BP PLC.
▪ The Russian Federation exported 217 Bcm mainly to Germany, Italy, and Turkey.
▪ Canada exported 73 Bcm exclusively to the USA.
▪ Norway exported 109 Bcm to several European locations; principally France, Germany, and the UK.
▪ The Netherlands exported 38 Bcm principally to Germany.
▪ Algeria exported 21 Bcm under the Mediterranean to Italy and Spain.
From this, it is unsurprising to see that the largest four importers of gas via pipelines
were:
▪ Germany
▪ USA
▪ Italy
▪ China
▪ Netherlands
109 Bcm
73 Bcm
54 Bcm
48 Bcm
41 Bcm
These main flows are shown graphically on Figure 7.6.
7.5.8
Liquefied natural gas (LNG)
Countries dependent on natural gas that (for geographical reasons) cannot receive it
via a pipeline must resort to receiving the commodity on a seaborne basis in the form of
LNG. This is not a competing product to natural gas, but merely an alternative method
of distribution. In liquefied form natural gas occupies 1/600 of the space it would in
gaseous form making it more viable to transport over longer distances. To create LNG,
it is cooled below its freezing point of −161 ∘ C(−260∘ F), and other constituents such
as oxygen and carbon dioxide are removed to leave a gas that is virtually pure methane.
Once converted it can be transported in specially adapted ships to LNG terminals in any
250
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
FIGURE 7.6 Movements of natural gas by pipeline. Billions of cubic metres.
Source: BP Statistical Review of World Energy 2020. BP PLC.
geographical location, where the conversion process is then reversed. However, converting natural gas to LNG is an expensive and complex process that requires a substantial
infrastructure, which is expensive and may take many years to construct.
Annual production of gas in LNG form is expected to increase from its current levels
particularly for countries with high consumption rates relative to their own domestic
supplies. The biggest importers of LNG are shown in Figure 7.7.
The largest exporters of LNG are shown in Figure 7.8.
These main flows are shown graphically on Figure 7.9. Although not immediately
obvious from the figures, there are certain countries (e.g. USA, Trinidad and Tobago,
Nigeria) that export small amounts of LNG to many countries.
7.6
NATURAL GAS PRICES
7.6.1
Natural gas price definitions
Before considering the factors that influence the price of natural gas, it is perhaps useful
to consider what is meant by the term ‘price’. From a retail perspective, the price of
natural gas may vary (due to price changes at the wholesale level), but may be influenced
by regulatory considerations. However, the price of natural gas at the wholesale level
will be influenced by a very different set of factors. In a highly monopolistic natural gas
251
Natural Gas
Japan
105
China
85
South Korea
56
India
33
France
23
Taiwan
23
Spain
22
18
United Kingdom
0
20
40
60
80
100
120
FIGURE 7.7 Major LNG importers. Billion cubic metres.
Source: BP Statistical Review of World Energy 2020. BP PLC.
Qatar
107
Australia
105
USA
48
Russian Federation
39
Malaysia
35
Nigeria
29
Indonesia
22
Algeria
17
Trinidad & Tobago
17
0
20
40
60
80
100
120
FIGURE 7.8 Major LNG exporters. Billion cubic metres.
Source: BP Statistical Review of World Energy 2020. BP PLC.
industry, contracts to buy or sell natural gas may well be long-term, bilateral, and fixed
price in nature. ‘Take or pay’ contracts might exist if a buyer doesn’t take delivery of the
agreed volume, they would have to pay a fee. Although aspects of these types of pricing
structures may persist in a re-regulated market, there is now a greater possibility that
the price will be based on fundamental demand and supply considerations. As a result,
a term structure of prices would gradually evolve along with the use of risk management
products.
As with any commodity there will be different prices for delivery of natural gas
for different maturities and for different locations. However, although there may be a
common set of delivery maturities, there will be different sources for reporting prices of
deals executed. To ensure the efficient settlement of a variety of contracts (e.g. physical
supply contracts or derivatives) the emergence of a single, credible, and accepted benchmark reference is key. For the energy markets there are a variety of different possible
252
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
FIGURE 7.9 Movements of LNG. Billion cubic metres.
Source: BP Statistical Review of World Energy 2020. BP PLC.
sources. For example, this index should be representative of market conditions at a particular time and that the market is liquid (defined as the ability to execute a transaction
without significantly moving the price). Some contracts may be settled against a price
quoted on a futures exchange such as ICE or NYMEX (which is part of the CME Group).
Examples of natural gas prices in different countries (UK, Holland, and the USA) are
shown in Figure 7.10.
Other popular price sources are privately owned, price reporting publications such
as the Heren European Spot Gas Markets (HESGM) published by ICIS and ‘Inside
FERC’ published by S&P Global Platts for the US market. The HESGM reports a variety
of different prices, which include:
▪ Monthly and daily price indices for a variety of locations (e.g. UK NBP, and continental hubs such as Zeebrugge),
▪ Quoted prices that cover natural gas for delivery day-ahead, weekend, working days
next week, the balance of the month, monthly, quarterly, and annually,
Inside FERC reports:
▪ Spot prices of natural gas delivered to a variety of pipelines.
▪ Spot prices for different market centres.
253
Natural Gas
12.00
10.00
8.00
6.00
4.00
2.00
0.00
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
UK NBP
NL TTF
US Henry Hub
FIGURE 7.10 Natural gas prices. USD per million Btu. NBP = National Balancing Point,
TTF = Title Transfer Facility. These are virtual pricing points through which all gas in that
country is deemed to flow.
Source: BP Statistical Review of World Energy 2020. BP PLC.
7.6.2
Oil indexation in the natural gas market
Where the price of natural gas is determined by the interaction of fundamental demand
and supply, this is referred to as ‘gas-on-gas’ pricing. The linkage between the price of
natural gas and crude oil (‘gas to oil pricing’) stems from a number of reasons and affects
both regulated and re-regulated markets. The linkage exists for a number of reasons:
▪ Both energy sources are often found together. By linking the price of natural gas
to oil energy, companies would not be incentivised to produce one source over
another.
▪ The early natural gas finds in Europe were valued based on alternative fuels, which
commercial and domestic consumers had used prior to this time. As a result, natural gas contracts for industrial consumers tended to be linked to the price of fuel
oil, power generation contracts to the price of fuel oil or coal, and domestic supply contracts to the price of gas oil. In some cases, contracts could also be linked
directly to the price of a particular crude oil (e.g. Brent).
▪ When natural gas became more of a substitute for oil, natural gas price rises became
indexed to the changes in the oil price to ensure that it remained attractively priced
and retained its market share.
▪ For some end users there is an element of substitutability between different fuel
sources, so relative pricing becomes important.
▪ Since oil-based products are mature liquid markets, they provide an attractive relative pricing source in geographical locations where natural gas markets are considered illiquid. This was particularly relevant when banks were lending to large
exploration and production projects. To ensure there were sufficient cash flows to
service the debt, this led to the development of long-term natural gas sales contracts
that were indexed to oil-based products.
254
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
One interesting illustration on the role of oil prices within a deregulated market
is the situation faced by the UK. Many large industrial consumers and power stations
can operate using either natural gas or oil-based products. As a result, the gas-oil price
link will have an indirect impact on the price of natural gas. Additionally, as the UK
becomes more reliant on overseas European imports, its main source of supply will be
from a variety of different countries within Europe. Since a significant proportion of
natural gas contracts in Europe are indexed to the price of crude oil, and if natural gas is
imported via the Interconnector, the price charged to the purchaser will be influenced
by the price of oil.
A report published by the European Commission in 2006 about competition in the
European Gas markets highlighted more detail about the nature of oil indexation in
the region. It noted that supply agreements were characterised by long-term contracts
of between 15–20 years in duration and generally included a pricing formula that was
indexed to a variety of different fuels. There is no single indexation formula that applies
across all markets but there may be common elements to the supply contracts in which
they are embedded:
▪ A predefined maturity, say, 10 or 20 years.
▪ A ‘take or pay’ clause that requires the buyer to take delivery of an agreed quantity
of gas or to pay for it even if it is not required. These contracts are also sometimes
termed ‘take and pay’ or ‘swing options’.
▪ The price of natural gas in these contracts is usually recalculated every one to three
months.
A hypothetical indexation formula might have the following characteristics. Let
us assume that a gas supplier was fixing their prices with a large industrial consumer
as of 1 September of a particular year. There will be a base price for the natural gas,
which may be adjusted by the average price of the chosen oil-based product (e.g. fuel
oil) that prevailed over a three-month period with a one-month lag. So, in this example,
the pricing window would extend from 1 May–1 August. The contract price would then
be valid for, say, a three-month period to 1 December. The effect is to link the price of
natural gas to the price of oil, but with the effect being both lagged and damped by the
averaging process. This indexation formula would be referred to as a ‘3-1-3’. The end
formula will normally include adjustments to ensure that the final price takes account
of the fact that the prices of the different oil products are expressed in units of volume
rather than units of energy, and may be traded in a different currency, and may have
different thermal properties.
The EC report suggested that about 75% of all contracts were indexed and the most
popular oil-based products were light fuel oil, gas oil, and heavy fuel oil. This figure
varied according to the region of natural gas production. In the UK, the figure was closer
to 20% with the main pricing basis being either market determined or one that was
linked to the rate of inflation. In the Netherlands, Norway, and Russia, the indexation
patterns were revealed that about 80% of production was indexed to heavy and light
fuel oils.
As a result, it would be reasonable to say that regulated markets are characterised
by longer-term contracts, with take or pay clauses and prices indexed to oil-based
Natural Gas
255
products. Deregulated markets would have a wider range of contract maturity from
within day contracts to longer-term contracts (but not as long as those seen in regulated markets). Prices are more likely to be indexed to the fundamental demand and
supply factors for natural gas. It would seem likely that the influence of crude oil and
the refined products will continue to exert an influence over EU pricing structures
since Netherlands, Norway, and Russia supply about 60% of Europe’s natural gas needs.
The advent of a re-regulated market will not necessarily signal the immediate end of
the influence of oil on natural gas prices given that contracts indexed to oil tend to be
very long term in nature.
7.6.3
Liquefied natural gas (LNG) prices
7.6.3.1
Overview of LNG contracts
LNG contracts are sometimes referred to as ‘sales and purchase agreements’ (SPA). They
represent a binding contract between buyer and seller for the purchase or sale of an
agreed amount of LNG for delivery over a specific period at an agreed price. The terms
of the contract would typically include:
▪ Duration – traditionally SPA contracts were very long term in nature (e.g. 20 years)
as the sellers needed to ensure sufficient cash flow to repay their capital investment. Although this term of contract still exists, there has been a gradual move
to shorter-term transactions combined with an increase in spot trading. Buyers
may also have infrastructure investments to finance and so may also be happy with
long-term contracts as they offer cash flow certainty.
▪ Volumes – although the contract may state an annual volume, there may be some
flexibility for the buyer. For example, it may allow them to take an increased amount
or to elect not to take part of their contracted delivery without activating any ‘take
or pay’ clause.
▪ Take or pay clause – this requires the buyer to pay for a minimum amount of LNG
whether they take delivery or not. Sometimes this clause may stipulate that a certain
percentage of their annual contract quantity must be taken; this might range from
85–100%.
▪ Destination clauses – in some LNG contracts there may be a clause that limits the
buyer’s ability to transfer the cargoes to another entity or that restricts the buyer
from nominating an alternative delivery terminal. From a buyer’s perspective this
clause could be restrictive, as having some degree of flexibility over the end destination would help them manage any ‘take or pay’ risks, particularly if the original
destination is oversupplied. Sellers were also concerned that buyers may be tempted
to arbitrage market prices by selling on a cargo to another buyer, which would
then mean it would become another source of supply. These destination clauses are
gradually becoming less common as certain jurisdictions may view them as being
anti-competitive.
▪ Payment terms – this will state whether the price is ‘free on board’ (FOB) in which
the buyer is responsible for the shipping costs or ‘cost, insurance, freight’ (CIF) in
which the seller is responsible for shipping costs.
256
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Pricing basis – Since there is no overarching global price for natural gas, this clause
describes how prices would be determined. It may be formula-based, referencing
crude oil, or based on an agreed gas index price such as the Henry Hub natural
gas future. The contract may also include some provision to renegotiate the pricing
terms during the life of the contract.
7.6.3.2
Asian LNG physical pricing
Asian LNG contracts are an interesting case study on how the pricing process has
evolved over time. In the very early days of the LNG market, contracts were often
simple fixed-price transactions. However, after the oil crises of the early 1970s, there
was a move towards using an index, which would capture the economics of higher
crude oil prices. Arguably, the most popular crude oil index formula is the Japan
Customs-cleared Crude (JCC), sometimes nicknamed the ‘Japanese Crude Cocktail’.
The typical structure of an LNG pricing contract that references the JCC
formula is:
Price of LNG = A × (Price of crude oil) + B
The price of LNG would be expressed in USD per MMBtu. ‘A’ was the degree
of indexation to oil prices, which could range from 11–15%. The magnitude of this
co-efficient was designed to reflect the degree of ‘oil/natural gas parity’. This is the LNG
price that would be equal to that of crude oil on a barrel equivalent basis. A co-efficient
of 0.1724 is full oil parity. If the agreed co-efficient was, say, 0.1485 this would suggest
that the contract had an 85% crude oil linkage (0.1485/0.172). So, a 1% increase in
the price of crude oil would mean that, all other things being equal, the price of LNG
would increase by 0.85%. The price of crude in the formula was the price of JCC, which
is based on the average price of crude oil imported into Japan. This could be based
on an average price over several previous months with an agreed time lag. ‘B’ was a
constant term that represented non-oil factors (e.g. shipping costs) and ensured that
LNG prices did not fall below a certain level.
Although this example illustrated how a link to the price of crude oil could be
established, since these LNG contracts are all individually negotiated, it is possible to
reference the price to other competing sources of energy such as a fuel oil, gas oil,
and coal.
Another pricing concept worthy of mention is the ‘S’ curve, a stylised example of
which is shown in Figure 7.11.
The basic idea is that there are different values for A and B, which depend on the
range of prices taken by the price of crude oil. In low crude oil price scenarios, there is
price support for LNG sellers, while the formula has a dampening effect on the value of
LNG at higher crude oil prices, which protects LNG buyers.
In more recent times, the market has gravitated towards using natural gas benchmarks with contracts referencing non-Asia indexes such as US Henry Hub or UK
National Balancing Point (NBP), given that there was no Asian natural gas pricing
point. As the market continues to grow there has been a move toward specific Asian
price indexes such as the S&P Global Platts Japan Korea Marker (JKM). The marker
257
Natural Gas
LNG contracted
price
Price support for buyers
LNG = A3 x JCC + B3
Price support for sellers
LNG = A1 x JCC + B1
LNG = A2 x JCC
Oil price
FIGURE 7.11 LNG ‘S’ curve.
reflects the spot market value of LNG cargoes delivered ‘ex-ship’ (DES3 ) into Japan,
South Korea, China, and Taiwan.
Figure 7.12 illustrates the movements in price between four natural gas price
indices from 2009 to 2019. The price correlation over this period between the JKM and
the various indices are:
0.97 − UK NBP
0.96 − NL TTF
0.36 − US HH
7.7
NATURAL GAS PRICE DRIVERS
7.7.1
Supply side price drivers
On the supply side of the price equation there are several key issues:
▪ Current levels of domestic production and proven reserves – for some countries, the
balance between the levels of domestic production and imports will have a big
impact on price. For example, the UK became a net importer of natural gas from
continental Europe in 2005, which introduced a new set of price dynamics such
3
DES means that the seller fulfills his obligation to deliver when the goods have been made available to the buyer on board the ship uncleared for import at the named port of destination. The
seller must bear all costs and risks involved in bringing the goods to the named port of destination.
258
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
18.00
16.00
14.00
12.00
10.00
8.00
6.00
4.00
2.00
0.00
2010
2011
2012
UK NBP
2013
2014
NL TTF
2015
2016
US Henry Hub
2017
2018
2019
JKM
FIGURE 7.12 Natural gas benchmark prices 2009–2019. USD per million BTU.
Source: BP Statistical Review of World Energy 2020. BP PLC.
as increased sensitivity to changes in the price of crude oil. If the amount of natural gas to be extracted falls below initial expectations, this may exert upward price
pressure.
▪ Production outages due to accidents or weather – for example, US natural gas production, which is very concentrated in the Gulf of Mexico, can be susceptible to weather
disruption (e.g. Hurricanes Katrina in 2005 and Harvey in 2017). Within this category, it would be appropriate to include the closing of fields for maintenance during
certain times of the year when demand is expected to be low.
▪ Available infrastructure – like many commodities there is a limit to the flexibility of
supply because infrastructure projects will require a long time to complete. Interconnectors are large international pipelines used to transport natural gas. If the
price of continental natural gas is higher than that in the UK there is an incentive
to deliver natural gas into Europe, while the opposite will hold true if the price of
UK natural gas is higher than that seen on the continent. The size and timing of
planned infrastructure projects such as storage facilities, new interconnectors, and
LNG terminals may influence prices for delivery of natural gas in the future.
One particular day, shortly after the opening of the Langeled Norwegian
pipeline to the UK, a combination of factors (warm weather, a major incoming
shipment of LNG, a shortage of storage capacity, and a large delivery to test
the interconnector) caused a surplus of supply. National Grid Gas reported that
344 million cubic metres were available compared with estimated demand of
234 million cubic metres. This led to suppliers paying buyers to take up the excess
with the price of gas reportedly falling to minus five pence per therm.
▪ Storage capacity – it is common for natural gas to be stored under and above ground
to absorb any change in the demand and supply balance. A lack of storage capacity
may cause prices to increase sharply if, for example, there is an unexpected spell of
cold weather.
Natural Gas
259
▪ Security of natural gas supplies – traditionally Western Europe has been able to
source its natural gas supplies from four key areas:
▪ The North Sea (from UK, Norwegian, and Dutch owned fields).
▪ Onshore European natural gas fields.
▪ Russia.
▪ Algeria.
Security of supply can be interpreted as either the availability or reliability of
natural gas supplies and is a key factor for countries that are reliant on imports
of natural gas, particularly where there is no diversification of supply. Security of
supply is enhanced by the existence of a diversified source of supply, adequate
transmission infrastructure, and sufficient storage facilities. Arguably, the converse could also apply, i.e. the security of demand; that is one country becomes
over-reliant on a single buyer and could suffer economically if that relationship
were to deteriorate.
▪ Proportion of natural gas that is exported – one characteristic of the natural gas market has been the increase in the trading of LNG. The more that is exported – perhaps
driven by higher prices in other regions – the lower the domestic supplies and, all
other things being equal, the higher the price. For example, at the time of the first
edition of this text, BP estimated that the US was a net importer of 109 billion cubic
metres (Bcm) of natural gas, but by 2019 they were net exporter of 123 Bcm, largely
due to the impact of shale gas.
▪ Concentration of suppliers – The Economist (2012) argues that liberalised markets
tend to hand power to big suppliers. A large supplier could manage its production
in such a way that spot prices could be influenced in a particular direction.
7.7.2
Demand side price drivers
The demand side of the market is driven by:
▪ Seasonal demand or weather – The main component of natural gas demand is
domestic for heating and cooking requirements and so a change in the weather
may lead to a change in demand.
▪ Degree of domestic competition between suppliers – As the choice of suppliers
increases the price of natural gas should fall.
▪ Electricity generation – Natural gas is seen as a more environmentally attractive fuel
to run electricity power stations compared to coal. The deregulation of the electricity market has further spurred the growth of natural gas. However, as technology
improves, the importance of renewable energy is likely to increase.
▪ Exports – Since the UK is now linked to the rest of Europe via a series of pipelines;
it is possible to export gas to other countries if there is an increase in demand.
▪ Storage injections – Typically natural gas is injected into storage in the summer
when demand is relatively lower.
▪ Relative prices of competing fuels – As the price of alternative fuels increase relative
to natural gas, there is greater incentive to produce and consume natural gas.
260
7.7.3
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
LNG price drivers
There are several more specific factors that influence the pricing of LNG contracts:
▪ The number of competing supply projects,
▪ Demand for LNG from other potential customers,
▪ Buyer’s willingness to take an equity stake in the project,
▪ Buyer’s flexibility around the destination of the shipment,
▪ The required volumes,
▪ Availability of LNG vessels,
▪ The level of reserves.
7.8
TRADING NATURAL GAS
One of the impacts of moving from a monopolistic structure characterised by long-term
fixed pricing agreements, is the move to shorter-term trading arrangements. The motivations for trading natural gas can be explained within the context of the risks faced by
different elements of the physical supply chain.
7.8.1
Motivations for trading natural gas
Producers will be concerned with maximising their revenues and meeting contractual
supply commitments. If they are unable to produce natural gas to meet their requirements, they may be forced to enter the market to purchase the amount they need from
a third party in order fulfill the delivery to their end client. Additionally, if they have
flexibility to increase their production they may be tempted to produce more if they
believe they will be able to sell it at a higher price. They may be faced with a number
of different supply contracts with different maturities that they may or may not be able
to meet from their existing production. Additionally, the contracts may be priced in a
variety of different ways (e.g. fixed, floating, oil indexation). This may motivate them to
enter into derivative contracts to mitigate the associated risk.
The flip side of this are the consumers of natural gas who will be looking to ensure
that they can supply sufficient gas at the most cost-efficient price. Like the producers,
they will be faced with a variety of market risks that they may wish to mitigate.
Since shippers are responsible for ensuring that their demand and supply commitments are in balance, they may be forced into the market to buy and sell natural gas.
The TSO will be responsible for ensuring that overall, the system is in balance and
so may need to buy and sell natural gas to achieve this.
With respect to storage, one of the popular strategies that are used by traders and
consuming utilities is to buy natural gas when it is cheap (e.g. in the summer months),
in anticipation that the revenues generated at its eventual point of sale will exceed all
associated costs.
Financial institutions (e.g. banks and trading houses) will offer risk management
solutions to those entities along the physical supply chain. As such they will take on
a particular risk that they will then seek to offset at a profit by trading with similar
institutions. For example, they may identify an opportunity to arbitrage the price of
natural in two different locations (e.g. UK NBP vs. Zeebrugge).
Natural Gas
7.8.2
261
Contract types
Natural gas contracts (physical or financial) can be traded either on an exchange or
over the counter basis. For example, within the UK there are two main organised
exchanges: the Intercontinental Exchange (ICE) and the On-the-Day Commodity
Market (OCM), both of which will result in physical delivery if the contracts are held
to maturity. The OCM market is designed to allow market participants to balance their
short-term demand and supply commitments. The OCM market also allows the system
operator to buy and sell natural gas for operational and safety reasons.
Physical contracts can be executed with a variety of different maturities:
▪ Same day (sometimes referred to as ‘within day’).
▪ Next day (variously called ‘prompt’ or ‘day ahead’).
▪ Future months (buying/selling of gas every day of the specified month).
▪ Future quarters (buying/selling of gas every day for the three months of a specific
quarter).
▪ Seasons – this could be the summer period (April–September) or winter
(October–March).
▪ Annual contracts – Either the calendar year (buying/selling gas every day for the
12 months covering January–December) or the gas year (October–September); buying/selling gas every day for a 12-month period covering October to the following
September.
7.8.3
Delivery points
Within any natural gas market, agreement must be reached as to where the transfer of
natural gas ownership will occur. As a result, several popular delivery locations have
evolved, that are either physical or virtual in nature.
There are many natural gas market hubs in the USA of which the most cited delivery
point is the Henry Hub in Louisiana, which is the physical meeting point for 12 intraand interstate pipelines. Traditionally, a significant proportion of natural gas would be
taken from various production points in the Gulf of Mexico and then fed to a number of
states through the East Coast, Midwest, and up to the Canadian border. However, the
advent of shale gas continues to reshape these flows. Although it is possible to buy and
sell gas in any location, the price may be quoted as a spread (referred to as the location
differential) to the Henry Hub figure.
In a similar vein, the Zeebrugge hub is a common delivery location in continental
Europe as it represents the physical junction of several international pipelines. If the
hub is physical in nature, then it is likely that there will be a range of other facilities
available to the market participants such as storage and treatment facilities.
In some markets there may be a notional or virtual settlement point, which covers
an entire network. In the UK there is a virtual delivery point referred to as the National
Balancing Point (NBP), which is a notional point within the national transmission system through which all natural gas is deemed to flow and about which all natural gas is
required to balance. The Dutch national transmission company has created something
like the UK, which is referred to as the Title Transfer Facility (TTF).
262
7.8.4
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Trading natural gas in the UK
Over-the-counter (OTC) trades broadly encompass four different types of transaction.
NBP spot and forward transactions, beach contracts, and interconnector trades. These
trades can vary in their maturity; for example, NBP trades are quoted from ‘within day’
out to three years. Each counterparty will undertake to make trade nominations to the
system operator as to the quantity of natural gas and the period of the performance to
which it relates. An NBP transaction could also be executed such that the buyer and
seller will trade the right of ownership of natural gas within the NTS. Since the natural gas is already within the NTS, the system operator will simply have to make an
adjustment to the accounts of the two shippers as to the quantity traded.
Beach trades are bilateral contracts where the purchase or sale will take place at
a specified natural gas terminal where it is brought ashore from a production facility.
These deals are executed for natural gas transactions outside of the NTS and can be
useful for a shipper who is looking to resolve an anticipated system imbalance. Interconnector trades for delivery at, say, Zeebrugge will operate in a similar fashion.
One of the key requirements of the Network Code is that shippers are required to
balance their daily injections with daily deliveries at the NBP. If actual natural gas usage
is not equal to the amount nominated, the shipper will be out of balance. The system
will be long natural gas if too much is delivered, or the withdrawals (‘offtake’) are less
than expected. The system will be short natural gas if the shipper did not deliver enough
or the offtake was more than expected. This imbalance can be anticipated by the TSO
by matching nominations made by different shippers to buy and sell natural gas. There
are several ways in which the system can be brought back into balance:
▪ Shippers can obtain or dispose of natural gas in the On-the-Day Commodity Market
(OCM).
▪ Shippers trade with producers outside of the national transmission system using
beach contracts.
▪ The system operator can buy or sell natural gas on the OCM if they believe an imbalance will occur and settle retrospectively with a particular shipper.
▪ The system operator or shippers use contracts to extract natural gas from storage.
▪ Certain clients have agreements whereby their supply can be interrupted.
7.8.5
On-the-Day Commodity Market (OCM)
The OCM was set up as part of the re-regulation process of the UK Natural gas industry
to help manage system imbalances. Access to this 24-hour electronic market is restricted
to the system operator and licensed shippers. Transactions can be executed between
these counterparties for either balancing purposes or for pure trading motives. This
regulated market currently has two different segments, the day-ahead market and the
OCM within-day market. The day-ahead market offers four different types of contract:
▪ Individual days
▪ Balance of the week
▪ Weekend strip
▪ Working days next week
Natural Gas
263
The OCM within day market offers three types of transactions. A locational trade
allows for the purchase or sale of natural gas at specific points on the NTS. A physical
trade allows for the purchase or sale of physical natural gas at the NBP. Title trades
represent a change of ownership of natural gas already at the NBP.
If National Grid Gas is required to buy extra natural gas to make the system balance,
an imbalance charge is subsequently made to the offending shipper at what is referred
to as the System Marginal Price (SMP). On the OCM the ‘SMP buy’ is the highest price
at which the system operator bought natural gas for the system, while ‘SMP sell’ is the
lowest price at which natural gas was sold out of the system. A related concept is the
System Average Price, which is a volume-weighted average of all buy and sells.
7.9
NATURAL GAS DERIVATIVES
7.9.1
Exchange traded futures contracts
A natural gas futures contract for the UK was first listed in January 1997. It can be used
for a variety of reasons:
▪ To help manage a market participant’s natural gas price exposure.
▪ As a means of taking exposure to the gas market without necessarily having an
underlying exposure. To avoid taking physical delivery, the investor would close
out the futures position prior to expiry.
▪ The futures market can also be used to obtain or dispose of supplies to the natural
gas market, as all the contracts will result in physical delivery at the NBP if held to
maturity.
7.9.1.1
Futures contract specification
Table 7.1 highlights the main features of the UK natural gas future traded on the ICE
and the Henry Hub natural gas futures traded on the NYMEX exchange, which is part of
the CME Group. It should be noted that each exchange trades several different types of
contract and the following table is designed to convey the main contract characteristics.
For the contracts traded on the ICE, if the contract is not closed out prior to its
expiry the counterparties will be expected to go to full delivery in equal measure on
each individual day that the contract covers. Month contracts are strips made up of
individual and consecutive calendar days, the total number of which is dependent on
the number of days in the month. These contracts will be listed up to 83 months in the
future. Quarter contracts are strips of three individual and consecutive calendar months.
However, the contracts are standardised to given time periods (e.g. January–March)
rather than being a negotiable set of consecutive months. Season contracts are strips of
six individual and consecutive contract months. Season contracts are specified as either
April–September or October–March. Annual contracts are strips of 12 individual and
consecutive contract months comprising January–December.
On NYMEX the most popular contract is for monthly settlement with physical
delivery at Henry Hub (HH). However, NYMEX (as well as the ICE) have recognised
264
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 7.1 Contract specifications for UK and US natural gas futures.
ICE
NYMEX
Contract size
Unit of trading
Five lots
One lot equals 1,000 therms of
natural gas per day
Quotation
The contract price is in sterling
and in pence per therm
0.01 pence per therm
One lot
One lot equals 10,000
million British thermal
units (MMBtu)
US dollars and cents per
MMBtu
UDS 0.001 (0.1c) per
MMBtu (USD 10.00 per
contract)
Monthly contracts listed for
the current year and the
next 12 calendar years
Minimum price
fluctuation
Contract
description
Last trading day
Delivery
▪ Annual contracts
▪ Season contracts
▪ Quarter contracts
▪ Month contracts
Trading will cease at the close of
business, two business days
prior to the first calendar day of
the delivery month, quarter,
season, or calendar year.
Contracts are for physical delivery
through the transfer of rights at
the UK National Balancing
Point. Delivery is made equally
throughout the delivery period.
Three business days prior to
the first calendar day of
the delivery month
Physical delivery at the
Sabine Line company
Henry Hub in Erath,
Louisiana
Source: ICE, www.theice.com; CME Group, www.cmegroup.com
that there are a number of other natural gas pricing points in the USA and Canada
and therefore offer a number of ‘basis’ contracts that are quoted as price differentials
between various pricing points and Henry Hub. For example, if a natural gas consumer
wanted to hedge the cost of natural gas for delivery at a different location than Henry
Hub, then the use of a HH future would leave them exposed to basis risk. That is, the
price they would pay for the delivery of their physical natural gas could be different than
that received under a Henry Hub delivered future. The issue of basis risk is considered
below in Section 7.9.2.2.
7.9.1.2
Applications of exchange traded futures
Hedging the purchase cost of natural gas
Let us assume that a UK industrial consumer has a commercial agreement with a producer to receive physical natural gas on a regular basis. The price agreement is based
on the prevailing ‘spot’ price at the time of delivery, as the producer is not prepared
to enter into a fixed price contract. The consumer decides to buy a single month natural gas future on the ICE for delivery in three months’ time. The price quoted for this
Natural Gas
265
future is 44 pence per therm. The consumer decides to close out the futures position
at the last possible moment to avoid physical delivery (the ICE future expires two business day prior to the start of the contract delivery period). We will assume that on the
close out date the price achieved for selling the future is 48p/therm. If we assume that
the futures and spot price have converged, the consumer takes delivery of the natural gas at the same price from his supplier, his cost will be 48p/therm minus the profit
on the future of 4p/therm. This gives a cost of 44p/therm, a sum equal to the original
futures price.
Futures calendar spreads for natural gas storage operators
For those companies who run storage facilities, they will either be storing natural gas on
behalf of another entity within the supply chain or will hold it for their own account. If
there is excess storage capacity, it may be possible to generate an arbitrage profit. Recall
from Chapter 2 on valuation that one approach to analysing the forward price of an
asset is the spot price plus the cost of carrying an underlying position for the duration
of the transaction. This cost of carry would be driven by such elements as the cost of
financing the purchase of the commodity and the cost of its storage. If the cost of storage
in a facility is less than the value embedded within an observed forward price it may be
possible for the storage operator to execute an arbitrage trade.
For example, let us assume that the storage operator has identified that they will
have excess capacity in one month’s time for a period of one month. Let us assume that
the price of natural gas for one- and two-month delivery is 50p/therm and 53p/therm
respectively. If they were to buy the natural gas for forward delivery in one month and
simultaneously sell it forward for delivery in two months, they would make a profit if
the associated costs of taking physical delivery in one month’s time and storing it are
less than the price differential earned from the futures trades. So, if the cost of storing the gas for one month is, say, 1p/therm, their profit would be 2p/therm (53-50-1).
This strategy is referred to as ‘going long the calendar spread’. That is, the trade is the
simultaneous purchase of a short-dated futures contract with the sale of a longer dated
futures contract.
The use of exchange traded futures by the various entities along the supply chain
for natural gas is summarised in Table 7.2.
Exchange for Physicals (EFPs)
An EFP represents a mechanism that will allow two entities to use the futures exchange
as means of fixing the price for a physical supply contract without using the exchange to
fulfill delivery. An example based on crude oil is covered in Section 6.8.2.5. Let us take
an example of a producer of natural gas who has agreed to supply 50,000 therms per
day to an end user for the month of August. We will assume that both sides are unable
to agree on the price terms and so agree to use the EFP mechanism. EFPs need to be
registered with the relevant exchange prior to the maturity of the particular futures
contract to which they relate. They agree to register this trade 15 business days prior to
the contract maturity. We will assume that five business days later the producer enters
266
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 7.2 Potential applications of exchange traded futures.
Participant
Nature of price exposure
Action
Producer
Shipper/Marketer
Exposed to falling prices
If long natural gas, exposed to
falling prices
If short natural gas, exposed to
rising prices
Similar to shippers
If long natural gas, exposed to
falling prices
If holding excess storage capacity
held
Exposed to rising prices
Sell futures
Sell futures
Traders
Storage
Consumer
Buy futures
Buy/sell futures
Sell futures
Buy short-dated futures, sell long
dated futures
Buy futures
the ICE futures market and sells 50 lots (1 lot = 1,000 therms) at the prevailing price of
44p/therm.
The consumer will do the opposite trade (i.e. sell 50 lots) but can choose the timing
of the trade to any point after the registration of the EFP but prior to maturity of the
future. There is no requirement for the two parties to execute the futures trades at the
same time. We will assume that the consumer executes the purchase leg the following
day when the futures price has risen to 46p/therm. Since neither party wants to use
the exchange as a mechanism for delivery of the natural gas, they agree to a nominal
price to close out their futures positions, which will also be used to settle the physical
contract. We will assume that they agree the futures close out price to be 45p/therm.
From the producer’s perspective they have:
▪ Sold the future at 44p/therm.
▪ Closed out futures position at 45p/therm.
▪ Delivered the natural gas to consumer at 45p/therm.
As a result, their net income on the transaction is 44p/therm. This is calculated
as the money received from the consumer for the physical natural gas (45p/therm) less
the 1p/therm loss on the futures transaction. From the perspective of the consumer
they will have bought the futures at 46p/therm and will close out at the agreed price of
45p/therm to make a 1p loss. They take delivery of the natural gas at the agreed close
out price of 45p/therm resulting in a net cost of 46p/therm. In both instances the price
paid or received for the natural gas was determined by the price at which the initial
futures position was executed.
7.9.2
Over-the-counter natural gas transactions
OTC contracts can be separated out into two main types, physical or financial. Physical natural gas transactions will involve the delivery of natural gas whereas financial
contracts will be cash-settled.
267
Natural Gas
7.9.2.1
Physically settled forward contracts
Fixed price forward contract
The following term sheet indicates the typical terms for a physical contract that is executed at a fixed price:
Seller:
Buyer:
Trade date:
Supply period:
Daily quantity:
Total quantity:
Price:
Bank/trading house
Client
November
06:00, 1 March to 06:00, 1 September (184 days)
25,000 therms
4,600,000 therms
Forty-five spot five five (45.55) pence per therm
In this contract the seller is responsible for delivering 4.6 million therms over a
six-month period in equal amounts per day (i.e. 25,000 therms per day). This type of
contract includes a clause that outlines the compensation that either party will have to
pay if they fail to perform as part of the contract. The market has adopted a common
set of terms and conditions to manage such a situation. Say that the client is unable to
receive the contracted amount. The seller would then have to dispose of the natural gas
in the marketplace. The compensation that either side of the contract would face for a
breach of contract is based on the system marginal price (SMP) traded on the OCM.
Floating price forward contract
Here the price is not fixed but will be based on an agreed index. A typical term sheet,
linked to the day-ahead price quoted by ICIS Heren, might look as follows:
Seller:
Buyer:
Transaction date:
Supply period:
Supply quantity:
Price:
Index price:
Bank/trading house
Client
November
06:00, 30 November to 06:00, 1 December
400,000 therms
A price expressed as a price per therm equal to the Index Price.
For delivery on a Business Day, that day’s price as stated in
pence per therm of natural gas for the ‘Day Ahead’. For
delivery on a day which is not a Business Day, that day’s
price as stated in pence per therm of natural gas for the
‘Weekend’; each as published in the column NBP under the
heading ‘NBP Price Assessment’ in the issue of ICIS Heren
European Spot Gas Markets, published on the Business Day
immediately preceding that day of delivery.
268
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
A US floating contract might contain the following terms:
Seller:
Buyer:
Transaction date:
Delivery period:
Contract quantity:
Price:
Index price:
Bank/trading house
Client
December
Each calendar day from and including 09:00 Central
Standard Time (CST) 1 January to and including 09:00
hours CST 31 January in the same calendar year.
6,000 MMBtu per calendar day in the delivery period.
A price expressed as a price per MMBtu equal to the Index
Price.
“NATURAL GAS LOUISIANA / SOUTHEAST HENRY
HUB” meaning the contract price will be a price stated in
USD per MMBtu of natural gas, calculated on the first
business day of the calendar month following each
month in the delivery period, published under the
heading ‘Final daily price survey - Platts locations (USD /
MMBTU)’: in the issue of ‘Gas Daily’ for such calendar
month.
Another variation of this type of structure is to link the price to an exchange traded
future. A hypothetical UK transaction linked to the price of the ICE future would have
the following terms:
Seller:
Buyer:
Transaction date:
Supply period:
Daily quantity:
Total supply quantity:
Contract price:
Bank/trading house
Client
March
06:00, 1 July to 06:00, 1 August (31 days)
20,000 therms
620,000 therms
‘Natural gas – ICE – Monthly index’ meaning the contract
price will be a price stated in pence per therm of natural
gas, calculated as the average of prices per therm of
natural gas published on the ICE on each day of the
calendar month preceding the prompt calendar for
delivery in the prompt calendar month.
In this contract a single price will apply for each daily delivery, but this will not be
agreed at the time of the trade. In this example, the price applied will be based on the
daily average of the July futures price, in the month prior to delivery.
269
Natural Gas
Option on physically settled fixed price forward
It is also possible to execute an option for physical delivery. For example, a client may
decide to buy a put option, which would give them the right to sell natural gas for delivery at the NBP. Indicative terms might look as follows:
Transaction date:
Seller:
Buyer:
Option style:
Option type:
Expiration:
Strike price:
Premium:
Total premium:
September
Bank/trading house
Client
European
Put
23 December
35p/therm
GBP 0.00675/therm
GBP 27,202.50
If the client exercises the put then bank/trading house will buy natural gas from
the client under a pre-agreed set of terms, an example of which is given below:
NBP trade
Buyer:
Seller:
Supply period:
Daily quantity:
Price per therm:
Bank/trading house
Client
06:00, 1 January to 06:00, 1 February
130,000 therms
35p
If the option is not exercised, then the seller simply collects the premium. The premium cost is calculated as 130,000 therms a day for 31 days at £0.00675 per therm.
Options on UK gas can trade with implied volatilities that range from 40–90% and are
often structured with Asian-style (i.e. averaging) payoffs.
7.9.2.2
Natural gas swaps
Cash settled transactions are attractive to either producers or consumers where they
wish to separate the physical delivery or receipt of the natural gas from the associated
price risk. Additionally, there may be traders and hedge funds that have no physical
capabilities but wish to take price exposure to the market without the need to trade in
the underlying instrument.
270
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Natural gas swaps can be used for a variety of purposes:
▪ To manage the basis risk of delivering natural gas at different physical locations.
This could be achieved by entering a swap where the cash flows are linked to prices
in different delivery locations.
▪ To transform the price risk of a physical purchase or sale (e.g. from a fixed to floating
exposure).
▪ As a means of exploiting a potential view on a rise in a natural gas price.
Vanilla natural gas swap
Take for example a hedge fund that decides to take exposure to the underlying price of
UK natural gas. It enters a natural gas swap, with a maturity of six months, starting one
year after it has been traded. The swap will involve an exchange of cash flows, where
one will be fixed, say, 70 pence/therm and the other will be based on an index price
taken from ICIS Heren. The floating index might be the daily, unweighted average of
day-ahead prices quoted by the same source. The floating side of the swap can also be
based on the arithmetical average of a futures price, such as the price of the front month
ICE UK natural gas futures.
The cash flows will be exchanged monthly and will be based on a notional
quantity of 30,000 therms per day and so with a 182-day period this would give a
total quantity of 5,460,000 therms. Settlement of cash flows would take place five
business days after the end of the monthly calculation period. By way of example let us
assume that at the end of one of the settlement months that the daily average of prices
quoted by ICIS Heren was 66.56785 pence (the accuracy of the calculations would be
stated in the deal confirmation and may extend to five decimal places). We will say
that the hedge fund is the payer of the fixed rate and receiver of floating and there are
31 days in the month. The cash flows to be exchanged are therefore:
Fixed cash flows
35,000 therms × 70p × 31 = GBP 759,500
Floating cash flows
35,000 therms × 66.56785p × 31 = GBP 722, 261.17
Net payment by the hedge fund is the difference between the two cash flows: GBP
37,238.83. By paying fixed on this transaction, the hedge fund is effectively taking the
view that the price of UK natural gas will, on average, be greater than 70 pence per
therm.
Basis swaps
Swaps can also be used to transform the price risk of buying or selling natural gas at a
particular location. Take the following US locational basis swap:
Natural Gas
Trade Date:
Effective date:
Termination date:
Notional quantity:
Calculation period:
Payment date:
Floating Amount A:
Floating Amount A:
Reference price:
Specified price:
Source:
Pricing date:
Floating amount B:
Floating Amount B:
Reference price:
Spread:
Specified price:
Source:
Pricing date:
271
16 January
1 October
31 March the following year
4,000 MMBtu per calendar day in each calculation period
Each consecutive calendar month, from and including the
effective date, to and including the termination date
In respect of each calculation period, the fifth business day
following the last pricing date in such calculation period
Payable by the client
NATURAL GAS – Houston Ship Channel (HSC)
The index price for spot delivery
S&P Global Platts ‘Gas Daily’
The first commodity business day during the calculation
period
Payable by bank
NATURAL GAS – HENRY HUB – CME plus spread
Minus USD 0.1500 per MMBtu
Settlement price for the calendar month and year
corresponding to the calculation period
NYMEX
The last commodity business day on which the relevant
futures contract is scheduled to trade on the exchange
There could be several motivations as to why this transaction could be executed.
Basis risk with respect to the natural gas market is where the price in one location moves
by a different amount to that elsewhere. This basis risk is often expressed as a spread
differential to the NYMEX future, which settles at Henry Hub. Say that natural gas to be
delivered for a certain date at the Houston Ship Channel is USD 2.85 MMBtu while the
CME futures contract for the same delivery is USD 3.00 MMBtu. The basis differential
for HSC would be minus USD 0.15 to the NYMEX futures contract.
Since the bank will be less interested in owning the underlying, they may wish to
take a view based on how the locational spread may evolve over time. Looking at the
direction of the cash flows above, the bank will make a profit if the basis differential
falls (‘tightens’). Imagine that the differential moves to NYMEX minus USD 0.10. The
bank could take an offsetting position where it receives NYMEX less the USD 0.10 and
pays HSC. The net result is that the HSC cash flows cancel out and the bank receives
more on the incoming CME cash flow than it has to pay out.
If the client has an underlying exposure to natural gas priced at HSC, the swap will
allow it to transform the exposure to that of Henry Hub. If the client is selling natural
gas, the price it receives will be paid away under the swap and it will be a net receiver
of Henry Hub minus USD 0.15.
272
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Index swaps
Another variation of the basis swap is an index swap. Here both swap cash flows are
floating but reference different maturities. One cash flow may reference a daily price
index, while the other could reference, say, a monthly index. So if a gas market consumer
has a physical contract to buy natural gas based on a monthly index, they could enter
such a swap where they agree to receive from the counterparty a cash flow based on
the monthly index and pay a daily index in return. Their exposure to movements in the
monthly index is neutralised and is replaced with a daily index price risk.
Swing swaps
In physical natural gas supply contracts, it may be possible for the two parties to agree to
a ‘swing option’, which is a right to take more or less of some agreed contractual amount.
Although these clauses may be somewhat involved in nature, in essence, the ability to
take delivery of an increased amount is effectively a call option on natural gas, while
the ability to reduce the amount delivered represents a put option.
A swing swap will be structured as a fixed-float transaction, but arguably the defining characteristic is the price index to which the floating side of the transaction is referenced. In some natural gas swaps, the floating price may reference a single monthly
index, but with a swing option it may reference the average of a series of daily prices.
To illustrate the concept, consider the following example, which is based on an idea in
Sturm (1997).
Suppose that at the start of a particular month, a US natural gas producer agrees
to sell all of their production to a consumer in particular region on a fixed price basis
of USD 2.00/MMBtu. Several days later market conditions change and prices for this
region start to rise. The producer can benefit from this price rise by means of a swing
swap. The transaction would work as follows:
Physical transaction
▪ Producer delivers the natural gas and receives a fixed price of USD 2.00/MMBtu
as per the terms of their commercial contract.
Swap transaction
▪ Producer pays a fixed price. Ideally, they would like to pay a fixed price equal to
USD 2.00/MMBtu such that it matches the cash flow received from the physical
transaction. However, they will pay the prevailing market price, which could be
different than this value.
▪ Producer receives a floating cash flow, which is based on an average of daily
prices from an agreed source.
▪ If it is assumed that the fixed price on the swap is traded at USD 2.00/MMBtu,
then the producer will profit from rising prices via the floating leg.
273
Natural Gas
Gas formula swap
In Section 7.6 it was noted that some natural gas participants paid for their natural gas
based on a formula indexed to crude oil. A gas formula swap would involve an exchange
of cash flows that would transform the formula into a single fixed price equivalent.
Consider the following stylised example, from the perspective of a EUR-denominated
natural gas consumer. It is assumed that the consumer is buying their supplies on a gas
formula basis.
Maturity:
Fixed leg payer:
Fixed leg:
Floating leg payer:
Floating leg:
One year
Client
Fixed price in EUR per kW-h
Bank
In EUR, [0.5057 + 0.0302 * Brent (6,0,3)] / EURUSD exchange
rate (1, −1,1)
The oil linkage is a constant, plus a proportion of the oil price, which is converted
into EUR. The value for Brent is the arithmetic average of, say, Dated Brent over the
previous six months (the 0 indicates there is no lag) and is applicable for the next
three months. The exchange rate is the prior month daily average of the EURUSD
exchange rate.
The net result is that the cash flow received from the bank finances the purchase
of their physical natural gas and on a net basis they end up paying a fixed price in EUR.
It may also be possible to increase the complexity of the swap by including additional
energy products such as gas oil.
Chooser swap
Another structure would be for a client to enter into a transaction where they receive,
under a swap, a cash flow linked to the price of natural gas less a ‘discount’ (e.g. the
NBP price minus 3 pence/therm). In return they would pay one of, say, three different
prices:
▪ The ‘day ahead’ NBP price.
▪ NYMEX HH futures price (converted into p/therm).
▪ ICE Brent crude oil price (converted into p/therm).
The choice of which energy price paid by the client is made by the structuring entity.
This example assumes that the client is buying physical natural gas on an NBP
basis and the fixed discounted cash flow receivable under the swap is used to finance
this purchase. However, this swap cash flow is reduced by a fixed amount, so it is not a
perfect offset to their physical cost. It would imply therefore that the floating swap cash
flow they must pay would offer them some benefit, i.e. whatever they are required to pay
it will offset the 3 pence discount on the fixed leg. As such, it suggests that the client
274
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Market price
Market price
Client
Purchase of
natural gas
Bank
Fixed price
FIGURE 7.13 Illustration of cash flow direction for a ‘double down’ swap.
is taking a view on the relative energy prices. The reason the client is able to receive
a discount to the NBP price is due to the fact that they have sold optionality to the
structuring entity. Rather than receive an explicit cash flow, the fixed price is adjusted
downward.
Double down swap
This type of swap is more aggressive in nature than the previous examples. It allows
the client to benefit from an improved swap rate but with an element of conditionality
attached. If the monthly price for natural gas is less than or equal to the agreed swap rate
the notional for that month doubles. An example term sheet for a consumer of natural
gas may look as follows:
Fixed rate payer:
Floating rate payer:
Maturity:
Settlement frequency:
Fixed price:
Floating price:
Double down feature:
Client
Bank
One year
Monthly
Expressed in natural gas units (e.g. USD/MMBtu or
GBP/therm)
Daily average of prevailing front month futures contract
If the floating price is less than or equal to the fixed
price, the notional for that month doubles
The basic outline of the transaction is shown in Figure 7.13.
In this example, the fixed rate on the swap would be lower than an equivalent ‘vanilla’ swap. This would reflect that the client has in effect sold a series of
monthly options to the bank. If the floating rate declines, the bank would exercise
the embedded option for that month in order to receive the fixed amount on twice the
notional. This should result in a net cash inflow, as the amount payable on the floating
leg will have decreased due to the fall in the price for natural gas.
275
Natural Gas
7.9.2.3
Option strategies
In terms of option-based strategies, the various vanilla option strategies presented elsewhere could be applied within the natural gas market (i.e. see the chapter on base
metals). The following transactions represent ideas that are perhaps more specific to
the natural gas market.
Swaptions
One product that links the swap market to the option market is an option on a swap
(a swaption). Some of the terminology for this product can be confusing, but it is common to describe them in terms of either being ‘payer’ or ‘receiver’ transactions. A buyer
of a receiver swaption has purchased the right to pay a fixed price (and therefore receive
a variable cash flow) in a swap. If the counterparties to the transaction do not wish to
enter into the swap at the point of exercise, it is possible to cash settle the transaction
at the point of expiry. In this case the terms of the swap agreed under the option are
marked to market at the swap rates that prevail at the time of exercise. The resulting
net present value of the cash flows is then paid to the counterparty for which the deal
has value.
The following hypothetical term sheet illustrates a ‘receiver’ swaption:
Option contract
Option buyer:
Option seller:
Option style:
Option type:
Option maturity:
Settlement:
Premium:
Swap contract
Swap maturity:
Payment frequency:
Total Notional amount:
Notional per month:
Fixed price payer:
Fixed price:
Floating price payer:
Floating reference price:
Floating price:
Client
Bank
European
Swaption
One month
Physical delivery into a swap
USD 5 mm
One year
Monthly for both fixed and floating
4,800,000 MMBtu
400,000 MMBtu
Bank
USD 2.50 per MMBtu
Client
Natural gas – Henry Hub – NYMEX
Final settlement price on last trading date for NYMEX
Henry Hub futures contract corresponding to the
calendar month of the swap settlement
276
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Another popular strategy not yet considered is a corridor, which is comprised of
a combination of options. Assume that there is a consumer of natural gas that wishes
to buy insurance against a steep rise in price. He can buy a call option with the strike
placed out-of-the money and then finance this by the sale of another option. In previous
examples such as the ‘min-max’ structure analysed in Chapter 5, the sold option had
been of the opposite type (i.e. an OTM put with a strike set to achieve zero premium;
Figure 5.8), but in a corridor the sold option is the same type but with a higher strike.
Let us assume that the consumer is faced with the following market conditions:
Trade date:
Effective date:
Maturity:
Notional amount:
Reference price:
Current futures price:
Pricing date:
Premium payment date:
Settlement date:
Settlement method:
1 March
1 April
30 April
1,750,000 MMBtu
Near month NYMEX future (i.e. April)
USD 6.74/MMBtu
The penultimate business day on which the futures
contract is scheduled to trade on the exchange
Five business days from trade date
Five business days after maturity date
Cash settled
Options on North American natural gas can trade with implied volatilities that
could range from 40–120%. An OTM European-style call option to cover the period 1–30
April priced with an implied volatility of 50% and a strike of USD 7.00 would cost USD
0.35 per MMBtu. Since the option has been written on a notional of 1,750,000 MMBtu,
this would give a premium cost USD 612,500. This could be partly financed with the
sale of a call option with a higher strike. The higher strike option will not completely
recoup the premium on the purchased option as this could only be achieved if both
options had exactly the same strike (a somewhat pointless exercise). Let us assume that
the consumer decides to sell an OTM call with the strike set at USD 7.50 – a level he does
not believe will trade. For simplicity we will assume that it has been priced at the same
level of implied volatility. The premium on the option would be USD 0.20 per MMBtu,
to give a total income of USD 350,000. The net cost to the trader is therefore USD 0.15
per MMBtu or USD 262,500 on the entire position. Settlement on the purchased call
option will be:
Max (0, Underlying price at expiry − strike price)
Settlement on the sold call option will follow the same logic except that the seller
will face increasing losses as the expiry price rises above the strike price.
The effect of the strategy will be that if the underlying price is below the strike of the
purchased call the consumer will buy the natural gas at the prevailing market price plus
the net premium cost. Between the two strikes the consumer will pay a fixed price
equal to the strike of the purchased call plus the net premium cost. Above the strike
of the sold call the net purchase price will increase, but still be less than the prevailing
277
Natural Gas
TABLE 7.3 At expiry payoff of consumer corridor strategy.
Underlying
price
(USD/
MMBtu)
Cost of
purchasing
the natural
gas
Settlement
on purchased
call option
(USD 7.00
strike)
Settlement
on sold
call option
(USD 7.50
strike)
Net
premium
cost
Net
purchase
cost of
natural gas
5.00
5.50
6.00
6.50
7.00
7.50
8.00
8.50
(5.00)
(5.50)
(6.00)
(6.50)
(7.00)
(7.50)
(8.00)
(8.50)
0.00
0.00
0.00
0.00
0.00
0.50
1.00
1.50
0.00
0.00
0.00
0.00
0.00
0.00
(0.50)
(1.00)
(0.15)
(0.15)
(0.15)
(0.15)
(0.15)
(0.15)
(0.15)
(0.15)
(5.15)
(5.65)
(6.15)
(6.65)
(7.15)
(7.15)
(7.65)
(8.15)
underlying price. This is shown in Table 7.3, which analyses the ‘at expiry’ situation
for the consumer, assuming both options are cash settled. We will assume that the consumer has agreed to buy the natural gas based on the same price at which the option
will be settled.
Bermudan option
A Bermudan-style option allows the buyer to exercise on a pre-agreed number of dates.
Consider the following example: an industrial consumer of natural gas buys a fixed
volume every month except for one month where they must close the plant for routine
maintenance. However, at the start of the year the exact date of the closure is unknown.
Under the terms of their normal commercial supply contract, they will buy natural gas
every month based on a daily averaging process. To hedge this price exposure, they have
been using cash-settled swaps, structured such that an offsetting floating payment is
received by the client who, in return, pays a fixed amount.
During the closure of the plant they will not have any underlying commitment to
buy the physical natural gas but will still be required to settle the swap. If the underlying price of natural gas has fallen, they will have a net settlement liability on the
swap. On the other hand, a rise in natural gas prices will result in a net cash inflow.
To hedge against a potential fall in prices, the consumer buys a Bermudan put option.
Once the closure dates are known, the option can be exercised if needed. The payout on
the Bermuda put option can be structured to be Asian-style to match the swap and the
underlying commercial contract:
Max (Strike − Average of Spot, 0)
The option strike could be set equal to the fixed rate on the swap and the cost of
the option could be built into the swap’s fixed price so the client is not faced with an
upfront premium cash flow. Since they are buying the option, the fixed rate payable on
the underlying swap will be higher than a vanilla equivalent. If the price of natural gas
278
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
falls then they will pay on the swap but will exercise the option and receive a compensating payout. If the price of natural gas increases, then they will receive on the swap
and allow the option to lapse.
Spread options
A spread option is an option that pays off based on the difference between the prices
of two underlying assets, relative to a pre-agreed strike. A call option will pay off if the
spread is greater than the strike, while a put option will pay off if the spread is lower
than the strike.
Call payoff = Max (Price of asset 1 − price of asset 2 − strike, 0)
Put payoff = Max (Strike − price of asset 1 + price of asset 2, 0)
There are several possible ways in which spread options could be applied within
natural gas markets:
▪ Some LNG contracts may reference traded natural gas prices with the possibility
that a particular cargo could be linked to the higher of the US Henry Hub future
or UK NBP price. Suppose a producer has negotiated such a clause into an LNG
agreement. This is a form of optionality, which could be ‘monetised’ by trading an
offsetting financial contract. The producer could sell a call on the spread between
the two gas prices structured as MAX (NBP – HH – strike, 0), priced in equivalent
terms (e.g. pence per therm). If the price of NBP rose relative to HH resulting in
an ITM option at maturity, they would be required to make payment to the call
buyer. However, under the terms of the physical contract they would receive the
higher NBP price, which could offset the losses. In addition, their income will have
increased, as they will have received the option premium.
▪ Suppose a producer of a commodity such as chlorine operates a gas-fired power
plant. If they faced competition from other producers in another location, their
competitive position may be harmed if the price of gas in their domestic location
increases relative to the overseas price. An option on the spread between the two
prices would help manage this risk.
▪ A producer of natural gas may have gas stored at different geographical locations
(e.g. UK and Holland). If they are concerned about an adverse change in the relative prices of the natural gas, then it may be possible to structure a spread option
between the two.
▪ Spread options could also be used to express views on transport and storage. Different locations will have different gas prices; therefore a spread option between the
prices at two different locations could be a view on transportation costs given that
the underlying commodity is homogeneous.
▪ Options on gas oil storage were examined in Chapter 6. Storage costs could be
thought of, therefore, as an option on gas prices at two different points in time.
Natural Gas
7.9.2.4
279
LNG derivatives
The pricing of physical LNG contracts was considered in Section 7.6.3.2. In some
respects, derivatives for this segment of the market will be similar in nature to those of
‘regular’ natural gas.
Where the physical LNG contract is referenced to oil prices then some form of
oil-linked derivative could be used to hedge the resultant exposure. With the US now a
major exporter of natural gas, pricing physical contracts referencing the Henry Hub
futures price has grown in popularity, which therefore provides producers and consumers with a benchmark index that can be used to hedge the resultant price exposure.
A trader who buys LNG in the US referencing Henry Hub prices may then decide
to sell it in either Europe (at NBP or TTF prices) or Asia (referencing JKM prices) and so
will have an exposure to the spread between the two prices. They will be able to make a
profit as long as the cost of transporting the gas between the two locations is less than the
differential between the prices. This is another example of spread exposure considered
earlier in this section.
Another popular ‘view driven’ transaction is to trade the spread between JKM and
TTF (or HH) using either futures or options.
CHAPTER
8
Electricity
‘One of the biggest surges in the electricity in the UK occurred at the end of an international
soccer match in 1990 when England lost to Germany. Demand soared by 2,800 megawatts,
which was equivalent to more than a million kettles being switched on as the English drowned
their sorrows with the answer to all the world’s problems – a cup of tea’.
—UK National Grid
8.1
WHAT IS ELECTRICITY?
All matter is made up of atoms, the core of which is called a nucleus. This nucleus is
made up of protons and neutrons and is surrounded by electrons. If electrons move
from one atom to another, a current of electricity is created.
In its raw form, electricity is useless to our daily needs. It is nearly always converted
into a different kind of energy at its end point of consumption. It is converted into heat
to cook food or boil water or into light and sound energy in order to watch TV or listen
to the radio.
Electricity is made by converting different forms of energy. Not much electricity
exists naturally in the environment and it is nigh on impossible to capture or store. On
Earth, there are many forms of stored energy, mainly in the form of plants and fossil
fuels, which can be converted into electrical energy. It is also possible to convert movement energy from wind and waves and light energy into electrical energy. Changing one
form of energy into another is termed ‘transformation’, while moving energy from one
location to another is called energy transfer.
The following flow diagram (Figure 8.1) summarises the path of energy from source
to useful form.
280
281
Electricity
Energy
source
(fuel, wind,
waves,
solar)
Transfer
Power
station
Transformation
Electrical
energy
Transfer
Homes,
factories,
hospitals
Transformation
Heat, light,
sound,
movement
FIGURE 8.1 Energy transfer and transformation: the path or energy from source to useful form.
8.1.1
Conversion of energy sources to electricity
Fuels
Inputs such as fossil fuels or nuclear fuel are used to boil water. The resulting steam is
used to drive a turbine, which in turn drives a generator, which produces the electricity.
The generator works by ‘electromagnetic induction’. This is the rotation of a wire within
a magnetic field, which induces an electric current. An old-fashioned bicycle dynamo
works in much the same way as it transforms movement energy into electrical energy,
which is then transferred to the bulb and transformed into light energy.
Solar energy
Heat energy is absorbed by solar panels and then used to heat water into steam.
The steam is then used to drive a turbine which produces electricity in a similar way
to above.
Wind, wave, hydro electric, and tidal energy
Movement energy is used directly to drive a generator. In this case there is no need to
boil water to turn a turbine.
8.1.2
Primary sources of energy
Electricity is sometimes referred to as a secondary source of energy as it must be created from some primary source. The primary fuel sources used in electricity generation
include:
▪ Natural gas and liquefied natural gas.
▪ Nuclear power.
▪ Coal.
▪ Oil.
▪ Water.
▪ Wind.
▪ Solar.
The choice of fuel will be dependent on how efficiently it generates electricity, the
cost of the fuel itself, and any waste products that may result from the process. The main
fuels used in electricity generation in the USA and UK are shown in Table 8.1, with the
figures used in the first edition of this text by way of comparison.
282
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 8.1 Fuels used in electricity generation.
Fuel source
Natural gas
Coal
Nuclear
Petroleum
Renewables
Other fuel sources
USA (2003)
USA (2018)
UK (2004)
UK (2016)
16%
51%
20%
3%
7%
2%
35%
27%
19%
1%
17%
1%
40%
33%
19%
1%
1%
3.5%
42%
9%
21%
1%
25%
3%
Source: EIA, Digest of UK Energy statistics; figures may not add up to 100% due to rounding differences.
The UK remains heavily reliant on natural gas as the primary source of fuel while
coal’s contribution to the energy mix has declined. The use of renewable energy has
increased significantly. US production is no longer dominated by coal and this could be
explained by the increased availability and relative cost effectiveness of ‘shale gas’.
One of the reasons why natural gas is a popular primary fuel is that it has a relatively
high thermal efficiency. Thermal efficiency relates the electrical energy produced to the
energy content of the input as in every step of electricity generation some energy is lost.
Burning fuel to boil water is particularly inefficient since a lot of energy is lost in the
form of heat. Take for example a fossil fuel fired power station, where a common sight
is the enormous clouds emerging from their towers. This is mainly steam taking energy
in the form of heat being released into the atmosphere.
For example, in a combined-cycle natural gas power station, the thermal efficiency
reaches about 47%. This means for every 100 units of natural gas used as a fuel source,
only 47 units are converted to usable electrical energy. The least thermally efficient
energy sources are coal (about 36%) and then nuclear power (about 38%).
There are some significant regional differences in the selection of the primary
fuel source. In Norway production of electricity is nearly 100% hydropower-based. In
years where there is excess rainfall the country can supply its surrounding neighbours
(Sweden, Denmark, and Finland) with a cheap source of electricity. However, when
the reservoirs are low, Norway will import electricity from its neighbours who are more
reliant on conventional primary fuel sources.
8.1.3
Commercial production of electricity
Although the general principles of electricity generation are universal, the actual design
of power stations varies. Three possible generation methods are:
▪ Combined Heat and Power (CHP),
▪ Combined Cycle Gas Turbine (CCGT),
▪ Open Cycle Gas Turbine (OCGT).
A CHP power station is one that typically uses natural gas to produce electricity but
also allows the steam and hot water that is produced at the same time to be captured for
other uses. A CCGT is an energy efficient gas turbine system where initially one turbine
Electricity
283
generates electricity from the gas produced during the combustion of the primary fuel
source. The hot gases are then passed through a boiler and the steam that is produced
drives a second turbine that generates electricity. An OCGT generates electricity from
gas produced during fuel combustion. This method of generation is not regarded as
being particularly efficient but OCGT plants can increase their load very quickly and,
hence, tend to be used as a reserve to respond to sharp increases in demand.
8.1.4
Measuring electricity
Consumption of electricity is measured in two ways. When reference is made to electrical power, the unit of measurement is a watt. Since a watt is a relatively small amount
of power it is more common to express it as a kilowatt (1,000 watts) or a megawatt
(1 million watts). The higher the watt or kilowatt rating of a particular electrical device
the more electricity it requires. If an electric light bulb is rated at 100 watts, it is describing an instantaneous quantity; it is consuming at any split second 100 watts. However,
a consumer of electricity will not want to use electricity over a split second but over a
period, which means the quantity has to be expressed as a rate. The amount of electricity
generated or used over a period of time is measured in watt-hours (Wh). This is determined by multiplying the number of watts used or generated by the number of hours.
For example, ten 100-watt light bulbs burning for one hour would consume 1,000 Wh.
When divided by 1,000 this is equal to one kilowatt hour (1 kWh) of electric energy. To
put this in some context the EIA estimates that the US residential sector used a total
of 1,462 billion kWh in 2018, with space cooling and heating accounting for the most
significant amounts (214 and 217 billion kWh respectively).
If someone wanted to have access to a generation resource, they might decide to buy
a power station with a capacity of 100 MW. However, if they wanted to use this amount
of electricity for one hour, they would buy 100 MWh. Other measurements of electrical
energy in ascending order are:
▪ Megawatt hours (MWh – a thousand kWh),
▪ Gigawatt hours (GWh – a thousand MWh),
▪ Terrawatt hours (TWh – a thousand GWh).
Another unit of measurement is the volt, which measures the force being used to
push electrons around a circuit. Technically this force is called potential, which is measured using the volt. Potential can be thought of as the steepness of a hill that electrons
run down. The steeper the hill (i.e. the higher the current), the faster they roll.
8.2
THE PHYSICAL SUPPLY CHAIN
Just like natural gas, regulation has played a major role in shaping markets in each
geographical location. A number of electricity markets have undergone substantial
deregulation followed by re-regulation. Issues of competition largely drove the original
restructuring of the electricity industry; however, since the first edition of this text
the consequences of climate change are now having a greater impact on the structure
284
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
of the industry. A common theme in the original process of deregulation was the
unbundling of the different functions that exist along the supply chain that were
traditionally performed by a single fully integrated utility. In the UK, which is regarded
as one of the most deregulated markets in the world, the result is that there is significant
competition in the generation and supply of electricity, but not in its transmission and
distribution.
The physical supply chain for electricity can be broken down into a number of different components:
▪ The production of electricity (generation),
▪ The transport of electricity on high voltage lines (transmission),
▪ Transportation on low voltage lines (distribution),
▪ The marketing of electricity to final consumers (supply),
▪ The buying and selling of electricity on wholesale markets.
The following descriptions outline the key functions of the main participants in the
physical supply chain of electricity.
Electricity generators – these will be the companies who build and run power stations that produce electricity. There may be a mixture of producers of different size
and varying degrees of integration. Some producers may operate a portfolio of different plants in different locations using different fuels and technology. At the other end
of the spectrum there may be independent power producers who do not have any transmission capabilities and do not sell on to the retail market. They may operate within a
defined area supplying electricity to wholesale buyers.
The transmission system operator (TSO) – having produced the electricity, it has
to be transmitted to the end user, usually by means of high voltage wires mounted
on overhead pylons. Transmission lines link the generators to the distributors. The
network of transmission lines is referred to collectively as the ‘grid’ with the TSO usually being responsible for ensuring its efficient and reliable operation. Typically, the
TSO will:
▪ Coordinate and schedule transmission transactions.
▪ Ensure that the demand for and the supply of electricity are balanced on an ongoing
basis. Since electricity cannot be stored, demand and supply must balance instantaneously, or the lights will go out.
▪ Manage generation in system emergencies.
▪ Manage generation reserves.
▪ Ensure that new transmission facilities are built when and where they are required.
▪ Coordinate transmission payments.
It is worth noting that TSO responsibilities vary by jurisdiction. In the UK, National
Grid owns the physical transmission network, but National Grid Electricity System
Operator (a separate legal entity) ensures that demand and supply are balanced in
real time.
Electricity distribution businesses – electricity cannot be delivered into a home
directly from the high voltage network so a series of substations transfer the electricity
Electricity
285
onto a lower voltage local network before its final delivery. These businesses may
sometimes be referred to as Distribution System Operators (DSO).
Suppliers – these are the companies that sell and bill customers for the electricity
that they use. They will make use of the distribution networks to supply energy to the
end consumer. In some cases, wholesale customers who have a significant demand may
agree on a supply contract directly with the producer.
End customers – they can be either domestic users or large and small business
customers.
Each of the participants in the supply chain will want to be compensated for the
service they provide and for the UK, the final price paid by the domestic consumer is
made up of:
▪ Wholesale costs (buying the energy in the wholesale market) – 32.32%,
▪ Network costs (costs of delivering the energy to a home) – 23.15%,
▪ Operating costs (billing, customer services, IT systems) – 17.34%,
▪ Government environmental and social obligation costs – 20.44%,
▪ Value added tax – 4.76%,
▪ Supplier pre-tax margin – 0.73%,
▪ Other direct costs – 1.25%.
Although this description of the supply chain is still valid, changes in technology
and climate change have impacted this structure. It is not the purpose of this book
to speculate on how the industry may evolve, but it is important to appreciate that
the structure is changing all of the time. For example, the production of electricity
is now impacted by the increased use of renewable sources such as solar and wind
power. Improved battery storage technology, homes fitted with ‘smart’ technology, the
increased use of domestic solar panels, and electric vehicles are reshaping the way the
industry will supply an end consumer.
8.3
MARKET STRUCTURE AND REGULATION
At a very high level two main electricity market structures have evolved: electricity pools
and bilateral markets. However, this is a simplified way of catergorising the different
structures and some markets may well not be fully described using these terms.
Electricity pools
Pools are sometimes characterised as being a centralised market, where competition
between generators is encouraged to achieve a fair price for buyers. In a compulsory
pool all of the generators will sell their output at a single price. In an auction-like process, typically overseen by a system operator, the generators submit how much they
are willing to generate and at what cost. The system operator will then ‘dispatch’ the
required energy either based on forecasted or actual demand.
In the auction-based process, prices are set for a single generation period and the
price from period to period may differ significantly. For example, if generating capacity
286
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
is suddenly lost, there could be a substantial increase in price until sufficient reserve
capacity can be brought online. Those entities supplying end customers indicate how
much energy they require and how much they are willing to pay; generators indicate
what they are willing to supply and the price they will charge. The auction process
sets the price for any period as the highest generating price submitted, such that all
demand is satisfied. The following is a simplified example of how such a system operates. Assume that in one single period of generation the demand to receive electricity is
1,000 MW, and there are four generators (I to IV). Each of the generators has the capacity to provide 250 MW but the prices at which they are willing to supply differ due to
the type of fuel used in the generation process. The prices offered by the generators are:
▪ Generator I: USD 60.00
▪ Generator II: USD 65.00
▪ Generator III: USD 70.00
▪ Generator IV:USD 75.00
In this case, supply and demand is matched in terms of volume, and the clearing
price (sometimes referred to as the system marginal price) for this period is USD 75.00.
This is the highest accepted price from all available generators such that demand
is equal to supply. All of the generators are paid this amount irrespective of how
much they bid in the auction process; this single price is also paid by those demanding
the electricity. In this example, all of the producers are ‘in merit’ as they will all
generate electricity. Had the demand for electricity been only 750 MW then generator
IV would not produce and would be deemed ‘out of merit’. The clearing price in that
case would be USD 70.00 set by generator III.
Bilateral markets
This structure requires both sellers and buyers to enter bilateral contracts for the sale
of electricity. This is the dominant model in the United Kingdom and a simplified
illustration of this structure is shown in Figure 8.2.
It is also possible for bilateral supply contracts to exist between generators
(see Figure 8.2) where one of the participants is unable to supply a contracted amount
of electricity. Once the transaction has been agreed, the participants must notify the
system operator by a specified time as to the net amount that will be bought or sold for
a given generation period. Unlike the pool system, it is now the seller of electricity who
will have to dispatch (i.e. generate) the agreed electricity based on the details outlined
in the bilateral contract. The system operator will then be required to ensure that the
system is always in balance and will have available specific tools that will allow them
to adjust the system accordingly.
Figure 8.2 also highlights the existence of organised exchanges. These will allow
participants to:
▪ Buy and sell physical electricity to balance demand and supply.
▪ Trade derivatives to manage the associated price risk.
Exchanges are considered in more detail in Section 8.5.
287
Electricity
PRODUCERS
Energy
Source
PRODUCERS
TRANSMISSION
DISTRIBUTION
RETAIL
CONSUMERS
INDUSTRIAL
CONSUMERS
BANKS,
TRADING
HOUSES,
BROKERS,
HEDGE FUNDS
SUPPLIERS
EXCHANGES
Flow of energy
Commercial supply contracts
Ad hoc physical supplies and / or risk management
FIGURE 8.2 Structure of bilateral electricity market.
Source: Author
8.3.1
The European Experience
Within the European Union it is important to distinguish between a regulation and a
directive. A regulation must be applied in its entirety within domestic law across the
European Union. A directive outlines a goal that all EU countries must achieve but it is
up to the individual countries to devise their own laws to achieve the objectives.
Within Europe, deregulation of the market began with the First Electricity Directive of 1996, which removed legal monopolies by allowing large electricity customers
the ability to choose their source of supply. It also set out the principles, whereby
vertically integrated companies were required to allow third parties access to their
transmission and distribution networks. It also started the process of unbundling (i.e.
legal separation) of the activities performed by these vertically integrated companies.
In effect, it introduced a distinction between the regulated part of the market (i.e.
the network operations) and the competitive elements (generation and supply).
Subsequent Electricity Directives and Regulations (2003 and 2009) developed many of
the themes of the earlier legislation. The overarching goal of EU legislation is based
on a ‘target design model’ aimed at creating interconnected European markets with
convergent prices.
At the time of writing, the aims of EU electricity regulation are to:
▪ Allow free movement of electricity throughout the EU market by means of cross
border trade.
▪ Increase competition and cooperation.
▪ Encourage the use of renewable energy to decarbonise the EU energy system.
▪ Improve consumer protection, information, and empowerment.
▪ Enhance the role of ACER (Agency for the Cooperation of Energy Regulators).
ACER was officially launched in 2011 because of the Third EU Energy Directive,
and its remit extends to both electricity and natural gas.
288
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Although it would be possible to fill an entire textbook on this subject, another
European trading regulation worthy of note is REMIT (Regulation of Energy Market
Integrity and Transparency). The regulation:
▪ Defines market abuse (e.g. market manipulation or insider trading).
▪ Prohibits market abuse.
▪ Requires effective and timely public disclosure of inside information by market
participants.
▪ Obliges participants to report suspicious transactions.
8.3.2
Overview of UK regulation
In March 2001, the New Electricity Trading Arrangements (NETA but pronounced as
‘neeta’) were introduced. NETA was designed to ensure that those entities wishing to
buy or sell electricity should be able to freely negotiate bi- or multilateral contracts either
on an OTC basis or on a recognised exchange. NETA included provisions for:
▪ A mechanism that allows for the balancing of demand and supply by the transmission system operator.
▪ Delivery of electricity to a single notional point avoiding the need for a multitude
of different delivery locations (the National Balancing Point).
▪ The development of short-term power exchanges that will allow market participants to fine tune their demand and supply requirements. For example, EPEX Spot
allows members to manage their within-day balancing requirements.
▪ OTC derivative markets that allow participants to hedge the exposures of varying
maturities.
▪ A settlement process to charge those entities that have not delivered or received
their contracted power.
In 2005, NETA was expanded to include Scotland and was renamed BETTA (British
Electricity Trading and Transmission Arrangements).
8.3.3
The American Experience
Although it is possible to generalise about the structure of an electricity supply chain,
the USA power industry has fragmented into several different markets with no single
overarching structure. In some areas of the USA, there are open, competitive, wholesale,
and retail markets while in others, a more regulated framework exists.
The supply of electricity in the USA has traditionally come from one of three
sources:
▪ Utility companies that may be owned by shareholders or the local municipality.
The utilities have responsibility for a specific geographical area and would typically
operate along all parts of the physical supply chain. Historically they may also have
supplied other household services such as natural gas, water, and telecommunications. If the entity is a municipal utility, one or more local governments may run it
depending on the area that it serves. Although utilities still exist in a deregulated
market, it is more likely that they will only be involved in one particular element,
such as retail distribution.
Electricity
289
▪ Rural co-ops set up to provide electricity to remote locations. This type of company
is owned by the customers it serves.
▪ Federally owned power agencies that may have been established to sell power
that is generated from infrastructure projects paid for by the government (e.g.
hydro-related). One example is the Tennessee Valley Authority, which was set-up
in 1933 to help the region develop during the Great Depression. One of its
original responsibilities was to provide electricity to homes and businesses in
the region.
The overall direction for the electricity industry is determined by the passage of
federal law. The Federal Energy Regulatory Commission (FERC) regulates the industry
at the wholesale level. Regulation at the state level encompasses areas such as vertically
integrated utilities and competition in the retail sector. Municipal utilities will either
be regulated at state or local government level, while electricity co-ops are regulated by
their own elected boards.
FERC’s main areas of responsibility include the activities of Independent System
Operators (ISOs) and Regional Transmission Organisations (RTOs), some aspects of
generation, and the wholesale trading of power. ISOs and RTOs are responsible for controlling, managing, and operating the bulk electric transmission grid. At the time of
writing there are seven ISOs/RTOs:
▪ California.
▪ ERCOT (Electric Reliability Council of Texas).
▪ Midcontinent.
▪ New England.
▪ New York.
▪ PJM.
▪ South West Power Pool.
North American Electric Reliability Corporation (NERC)
The Energy Policy Act of 2005 called for the establishment of an Electric Reliability
Organisation and a year later the designation was awarded to NERC, which was originally established in 1968 as a response to a serious blackout a few years earlier. NERC is
responsible for ensuring the reliability, security, and adequacy of the bulk power system
in the USA, Canada, and parts of Mexico. Their coverage includes the high voltage transmission system but does not extend to the distribution system that delivers electricity
into the home.
NERC defines reliability in terms of:
▪ Adequacy – the ability of the system to always supply the aggregate demand requirements of the end users.
▪ Reliability – the ability of the system to withstand sudden disturbances. In recent
times this has expanded to include the security of the system from physical or cyber
attacks.
290
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
NERC is responsible for designing reliability standards for all the US markets but
delegates some of its compliance activities to a series of partners known as regional
reliability councils, which are in turn made up of members from each segment of the
electricity industry. The regional reliability councils are:
▪ Western Electricity Coordinating Council (WECC).
▪ Midwest Reliability Organization (MRO).
▪ Southwest Power Pool (SPP).
▪ Electricity Reliability Council of Texas Inc. (ERCOT).
▪ Northeast Power Coordinating Council (NPCC).
▪ Reliability First Corporation (RFC).
▪ Southeastern Electric Reliability Council (SERC).
▪ Florida Reliability Coordinating Council (FRCC).
One of the consequences of having several different regulatory bodies is that it
creates a complex set of rules and regulations.
8.3.4
Wholesale markets in the USA
Perhaps due to the way that the US market was deregulated and re-regulated a patch
work of different markets has subsequently evolved. The US federal government has
the power to oversee interstate markets, but individual states oversee the utilities that
serve retail customers. So, because federal regulators did not dictate the structure for all
power markets, they preferred instead to provide broad guidelines, so each region was
free to structure its own market.
Although the US markets have evolved at different rates throughout the country,
once the traditional model of a single entity operating along the entire supply chain is
subject to deregulation, the development of independently owned generation emerges
to provide the incumbent utility, which at this stage of reform is the single buyer of
power, with more competition. Typically, the next stage of development in the market
allows large consumers to choose their source of supply. This gives rise to an Independent System Operator (ISO) that manages the transmission of power over a given
area. As competition increases, electric marketers emerge, buying and selling electricity
between wholesale participants. The final stage of development occurs when the retail
buyer is free to choose their supply of electricity.
As an illustration of how deregulation has impacted the US market, according to
the EIA’s website, in 2017 electricity sales were made by four major types of entity:
▪ Investor-owned utilities: 51%.
▪ Power marketers: 22% (e.g. entities that may not own generating facilities such as
financial intermediaries but can buy and sell electricity).
▪ Federal, state, and local utilities: 14%.
▪ Electric cooperatives: 11%.
▪ Other: 2%.
Electricity
8.4
291
PRICE DRIVERS OF ELECTRICITY
Introduction
Electricity must be produced when it is demanded. As a result, the spot price for
electricity is very dependent on its physical infrastructure. If there are problems with
physical delivery such as transmission constraints this could lead to volatile prices.
However, since it is impossible to store electricity, a high price today does not mean a
high price tomorrow. If demand exceeds supply of electricity, then the lights will go out
(a blackout). However, there may be instances where they may be a momentary excess
of demand over supply, which may cause the lights to dim or flicker. This is referred to
as a brownout.
The four main users of electricity are:
▪ Domestic (e.g. lighting and appliances),
▪ Industrial (e.g. manufacturing),
▪ Commercial (e.g. office buildings),
▪ Transportation (e.g. railways).
Electricity markets are essentially local markets given the fact that generation and
supply is generally limited to a certain geographical location. However, the markets are
subject to global influences as the primary fuel sources may need to be imported and
the cost of carbon emissions needs to be considered.
Load
Demand for electricity is related to the concept of load, which is defined as the amount
of power carried by a system or the amount of power consumed by an electrical device
at a specified time. This demand will vary considerably according to the time of day,
the day of the week, the season of the year and the climate. Generators will therefore
create a load profile or load shape that describes the pattern of electricity demand over
a given period.
Generating companies may operate different types of power stations according to
different types of demand or load. A baseload-generating unit is used to generate power
at a flat rate around the clock. Baseload is the minimum amount of electric power delivered or required over a period at a steady state. A peak load generating unit is used to
meet the requirement during periods of greatest demand on the system. Intermediate
load generating units meet systems requirements that are greater than baseloads but
less than peak loads.
When constructing load curves, (a representation of demand in a particular generation period) account must be taken of:
▪ The type of end user (e.g. residential, commercial, and industrial) and how they use
electricity.
▪ The time of day.
292
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Time of year.
▪ Geographic region. For example, in the USA, the demand for electricity peaks in
the summer except for the northern states and Florida.
Demand and supply – an overview
Figure 8.3 shows the global generation of electricity over the period 1985–2019.
Perhaps somewhat unsurprisingly, overall generation has more than doubled in the
period. Although generation in regions such as the USA and Europe has continued
to increase, the most noticeable change is the Asia Pacific region, dominated by the
growth in generation in China. In 1985, the country generated 411 TWh and by 2019
this had risen to 7,503 TWh. Even between the publication of the first and second
editions of this book (about 11 years) Chinese electricity generation has increased by
about 300%.
Since electricity is a secondary source of power, it is instructive to look at the different sources of primary fuel used to generate electricity. Figure 8.4 analyses the main
geographical regions and looks at their reliance on primary fuels in relative percentage
terms. Some key themes emerge:
▪ Coal still plays an important role in power generation, especially in Asia Pacific.
▪ A relatively large proportion of electricity in Central and South America is generated by hydro.
▪ Regions with large supplies of natural gas such as North America, the CIS, and the
Middle East use this commodity as their primary fuel input.
▪ Despite the current development of renewable technology, the application within
the power markets is still limited, with Europe being a leader in relative terms.
30,000
25,000
Africa
Middle East
S & C America
CIS
20,000
Europe
15,000
N America
10,000
Asia Pacific
5,000
19
8
19 5
8
19 6
8
19 7
8
19 8
8
19 9
9
19 0
9
19 1
9
19 2
9
19 3
9
19 4
9
19 5
9
19 6
9
19 7
9
19 8
9
20 9
00
20
0
20 1
0
20 2
03
20
0
20 4
0
20 5
0
20 6
0
20 7
0
20 8
09
20
1
20 0
1
20 1
1
20 2
1
20 3
1
20 4
1
20 5
16
20
17
0
FIGURE 8.3 Global electricity generation (Terawatt-hours) 1985–2019.
Source: BP Statistical Review of World Energy 2020. BP PLC.
293
Electricity
100%
90%
80%
70%
Other
Renewables
Hydro electric
Nuclear energy
Coal
Natural Gas
Oil
60%
50%
40%
30%
20%
10%
0%
North
America
S. & Cent.
America
Europe
CIS
Middle East
Africa
Asia Pacific
FIGURE 8.4 Global electricity generation by fuel type (2019).
Source: BP Statistical Review of World Energy 2020. BP PLC.
100%
90%
Imports, 4.30%
80%
Gas, 26%
Imports, 7.22%
Gas, 42.05%
70%
60%
50%
Coal, 43%
Hydro, 1.31%
Solar, 4.08%
40%
Wind, 17.08%
30%
20%
10%
0%
Coal, 2.18%
Hydro, 1%
Solar, 0.40%
Wind, 3.90%
Biomass, 0.70%
Nuclear, 20.70%
2012
Biomass, 6.63%
Nuclear, 19.44%
2019
FIGURE 8.5 Fuels used in the production of electricity in the UK. 2012 vs. 2019.
Source: http://www.mygridgb.co.uk/historicaldata/
Figure 8.5 looks at how the mix of fuels used in the production of electricity in the
UK has evolved over the period 2012–2019. Some of the key points are:
▪ The reliance on coal has decreased dramatically.
▪ The use of natural gas has increased.
▪ There is a greater reliance on ‘renewable’ energy sources such as biomass, wind,
and solar.
294
8.4.1
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Demand for electricity
Economic activity
Like crude oil, an increase in economic activity will tend to lead to an increase in the
use of electricity.
Weather
Using the USA as an example, seasonal demand for electricity will increase during
the summer as more households switch on their air conditioning units. Elements of
weather that affect the demand for electricity will include the absolute temperature,
the level of humidity, the amount of cloud cover, and expected rainfall. Some analysts
will refer to Heating or Cooling Degree Days. Heating Degree Days (HDD) are defined
relative to the outdoor temperature and what is a comfortable room temperature (e.g.
21∘ C or 70∘ F). The colder the weather, the higher is the number of HDDs as it suggests a greater energy demand to heat a building. Cooling Degree Days (CDD) work
in the opposite direction; when the weather is warmer (i.e. above 21∘ C/70∘ F) then the
number of CDDs is positive.
Human activity
The time of day will determine if the demand for power is heavy or light. Normal working hours are typically defined as peak time while off peak covers evenings, nights, and
early mornings. In the same vein, the day of the week will also impact demand. When
buying or selling power, weekends can be traded separately and may be classified as a
lower usage period. Extending the argument further, power can be traded in ‘blocks’ of
months to represent the different seasons, which in the electricity market are broadly
defined as winter (October–March) and summer (April–September).
End user
Suppliers will also differentiate between retail and industrial customers. Typically, retail
customers will be greater users of electricity at the weekend while industrial users will
represent the greatest source of demand during the week.
Special Events
There may also be odd events that will cause a sudden surge. For example, in the UK the
system operator must deal with what they refer to as ‘TV pick up’. This a short but sometimes dramatic increase in demand that occurs at the end of certain TV programmes
such as soap operas and sporting events. In their preparation for major sporting events,
the demand forecasters will collect a lot of information to develop an accurate picture of
demand. This included the timing of the match and whether it would coincide with sunset, if extra time were a possibility, and what alternative TV options would be scheduled
for after the match.
Electricity
295
Efficiency improvements
As technology improves, several household appliances have improved in terms of their
efficiency of consumption, leading to a decrease in their use of electricity.
Electric Vehicles
At the time of writing, the impact on the electricity market of an increase in demand
for electric vehicles (EV) is not clear. One argument is that demand for EVs will
significantly increase when the economics of owning the vehicle are superior to
the traditional internal combustion engine. One part of this issue is the creation of the
appropriate recharging infrastructure, which represents a significant investment. One
useful model of assessing the impact of EVs on the electricity system is suggested by
Massey (2018).
Number of EVs
×
EV usage (e.g. kilometres driven)
×
EV efficiency (kWh∕km)
= electricity used (kWh)
+ grid losses of 7–8% (getting the energy from production to consumption)
= required electricity generation in kWh
The use of EV poses a number of questions and challenges to the system such as
whether the system would be able to cope with a large number of people returning home
from work and immediately charging their cars during traditional peak load periods.
It may require the introduction of ‘managed’ charging where EV owners may have to
charge their vehicle during the night when demand is normally lower.
8.4.2
Supply of electricity
Physical capacity
There are several different factors that could be included in this category:
▪ The type of available generation (i.e. coal or natural gas), which will determine the
cost of generation and have an influence on the price paid for electricity in any
period.
▪ The operating ‘health’ of available generation, which might be reflected in the
amount of required maintenance on plants or outright mechanical failure.
▪ Limits on the physical capacity of electricity transmission that could restrict the
amount of power imported into or exported from a particular geographic region.
296
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
One relationship that many traders will monitor is the way in which the reserve
margin of a particular market changes. This is defined as the actual amount of available
generation capacity relative to peak demand and is normally expressed as a percentage.
For example, in the USA, FERC (Federal Energy Regulatory Commission) suggests that
capacity margins should on average be in the region of 15–17%. There is an inverse
relationship between the reserve margin and the wholesale price of electricity.
Interconnections
The term can be used to refer to individual transmission lines that link different markets
in one geographical location (e.g. the USA) or more commonly to the overall capacity
provided by the various lines (sometimes undersea cables) that run between two countries. The UK has interconnectors with France, Netherlands, Northern Ireland, and the
Republic of Ireland. In theory a difference in price between the two countries should
result in power being directed to the highest price location. The UK-France interconnector is approximately 70 km in length with 45 km of cable underwater. The capacity
of such interconnectors affects the ability of the country or region to import and export
electricity.
Fuel prices
Like all traded products, the balance between supply and demand will determine the
market price. However, it could be argued that in the short-term, demand for electricity is somewhat inelastic in its response to changes in price. For example, a domestic
user will switch on a light without much consideration for the actual cost. So if the
short-term price were to increase significantly for one generating period, it is unlikely
that this would have a material effect on the end user as most domestic and a significant proportion of industrial and commercial consumption is based on tariffs that do
not change rapidly.
As a result, it is the activities of the generating companies that determine the
short-term price for electricity. Suppliers seeking to buy electricity will seek out the
lowest cost source of generation such as wind and hydro. Once generation capacity
in this sector is reached, buyers will then purchase their supplies from the next
cheapest source, which will typically be nuclear powered. After this, electricity will be
provided by installations using either coal or natural gas as their primary fuel input.
Consequently, the price of electricity will be driven by the price of the fuel used at the
point where demand equals supply; the so-called ‘marginal fuel’.
Typically coal and gas generators are the marginal sources of electricity generation, therefore the demand and supply dynamics of these markets will have a significant
impact on the market price for electricity. In addition, and depending on which market
is being analysed, the cost of emitting carbon may also influence the price of electricity.
▪ Electricity and natural gas – An increase in the price of electricity should lead to
an increase in the price of natural gas. However, the same relationship may not
hold the other way around. At any point in time, generation capacity that uses the
Electricity
297
cheapest marginal fuel will be running near capacity. However, demand is unlikely
to be met by generation powered by a single fuel source, and so the more efficient
plants that use the alternate marginal fuel will be switched on. It will be the price
of the fuel input used by these generators that will determine the market price of
electricity. If natural gas prices are low relative to coal, installations using this fuel
input will always be on. To supply the marginal demand, coal-fired installations
will be used. If the price of gas rises but is still below coal, the increase is unlikely
to have an impact on the price of electricity.
Conversely if gas is more expensive than coal, a rise in its price will have a
greater impact on the price of electricity.
▪ Electricity and coal – The effect of coal on the price of electricity may be country
specific. For a country that imports coal from overseas, a rise in the price of electricity is unlikely to have a significant impact on its price since it is being purchased
in a global market subject to a multitude of different price drivers.
Like natural gas, the impact of an increase in the price of coal will depend on
whether it is the marginal fuel. If it is the marginal fuel, then an increase in the price of
coal will lead to an increase in the price of electricity. However, if coal is comparatively
cheap compared to natural gas then an increase in the price of coal will have a weaker
effect on the price of electricity.
There is also a positive correlation between coal and natural gas in that a
rise in coal prices will cause generators to demand more natural gas, pushing up
its price.
Electricity and renewable sources of energy
Typically, natural gas is the main source of energy used to produce electricity. One of
the UK regulators, Ofgem reported that in 2018 that although 32.8% of generation used
natural gas, wind power and solar made up 17.5% (2019).
Participants’ bidding behaviour
In markets where electricity is priced using an auction process, the bidding behaviour
of the participants may influence the clearing price.
8.4.3
Factors influencing spot and forward prices
The previous list of price drivers was presented without any context of time. Like many
commodity markets, participants will be buying and selling electricity for a variety of
different maturities resulting in a forward curve. Like any forward curve, it represents
the clearing price where demand equals supply for different future maturities. Like
many commodity markets it will likely display backwardation, albeit on a seasonal basis.
Power prices in the UK and Europe will be higher for winter delivery than those for
summer. However, traditional forward pricing relationships cannot be applied to electricity markets due to the inability to borrow, lend, or store electricity. Hence the idea of
298
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
a forward price being the expected future price less a risk premium would be perhaps a
more valid model (see Chapter 2 for more details).
In their report on the energy sector, the EU Commission (2006) noted that market
participants rated the following factors as being key price determinants for different
maturities:
Shorter-dated maturities
▪ Plant availability
▪ Input fuel prices
▪ Precipitation
▪ Wind speed
▪ Interconnector availability
▪ Temperature
▪ Price of carbon emissions
Longer-dated maturities
▪ Forward fuel prices
▪ New generation capacity or retirement of existing capacity
▪ Water reservoir levels
▪ Weather trends
▪ Interconnector capacity
▪ Price of carbon emissions
▪ Economic growth
The report also argued that a risk premium would be embedded within the
forward price to reflect the value to participants of the certainty of cash flow that a forward contract would offer over an unknown future spot price. This component could
be either a premium or a discount, but the report argues that, in practice, it is more
likely to be a premium. This is consistent with one of the pricing frameworks outlined
in Chapter 2.
8.4.4
Negative prices
One aspect seen in some energy (and agricultural) markets is the concept of negative prices. One example in the UK natural gas market was cited in Chapter 7. With
respect to electricity, intraday prices in Germany fell to −EUR 500/MWh on one day
in 2012 (Risk, 2016). One explanation for negative prices was given by the European
Commission (2019):
‘Negative hourly prices usually appear when demand for electricity is lower
than expected and when variable renewable generation is abundant, combined with
ongoing relatively non-flexible large baseload power generation (e.g. nuclear and
coal-fired plants). In such cases, conventional power plants begin to offer their output
for a negative price in an effort to avoid switching the unit off and having to go through
the costly and high maintenance operation of restarting the facility when they are
required again.’
Electricity
299
Other reasons cited for negative prices include:
▪ Transmission constraints that would restrict the movement of the excess power to
other parts of the grid.
▪ Government subsidies to renewable generators. This means that the price of
electricity can be negative but with a subsidy payment from the government the
generator could still possibly make a profit.
Specific examples of negative prices in the US include: (Risk, 2016)
▪ Hydroelectric plants in the North Western States when heavy spring rains force the
plants to keep running to avoid flooding.
▪ Solar and wind power combined with transmission outages in California.
▪ Negative prices in the PJM region, particularly North Illinois because of strong
winds in the Midwest. This would typically occur at night when demand would
be low, and the winds would be at their most powerful.
▪ Excess wind power in west Texas, prior to the development of new transmission
capability.
Negative prices were also seen in the United Kingdom electricity markets in 2019
(Watt Logic, 2020).
▪ In March of that year the temperature was unseasonably warm and so demand fell.
However, solar generation increased significantly leading to negative prices which
fell as low as −GBP 70.24/MWh for about six hours.
▪ Stormy weather in December saw high wind output that triggered negative prices.
By midnight on the 7th, wind generation accounted for just under 48% of generation. Prices were negative from around 10:45 pm on 7 December to 12:45pm on 8
December.
8.4.5
Spark and dark spreads
Since the main sources of fuel input to produce electricity are traditionally natural gas
and coal, a producer of electricity is caught between two markets. They run the risk
that the price of their input exceeds the price of their output. This is captured in two
key price relationships.
The spark spread is measured as the price of electricity minus the price of natural
gas. This effectively measures the relationship between what is produced (electricity)
and the main fuel input (natural gas). It is also possible to measure the ‘dark spread’,
which relates the price received for electricity to the cost of producing it by burning
coal. Although it would appear to be a simple calculation, the different energy sources
will be measured in different units. The convention is to express the spread in currency
units per MWh, which means that in calculating the spark spread, natural gas, which
will be expressed in therms (for the UK market) or British thermal units (for the US
market), will need to be converted into an equivalent MWh value. In addition, the thermal efficiency of the input needs to be considered. Recall from an earlier section that
thermal efficiency relates the electrical energy produced to the energy content of the
300
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
input. When calculating the spark spread, the market convention is to use a thermal
efficiency of 49.13%. That means that for every 100 units of natural gas that are used in
production, only 49.13 units are converted into usable electricity. To calculate the dark
spread, a thermal efficiency of 36% is applied.
In simple terms, the spark spread can be calculated as:
Raw spark spread = Price of power − (price of input fuel∕thermal efficiency)
However, with the advent of the emissions trading the raw spark spread is adjusted
to include the cost of emitting carbon into the atmosphere as represented by the cost of
a carbon credit.
Adjusted spark spread = raw spark spread − (CO2 adjustment ∗ CO2 price)
This adjusted spark spread is sometimes referred to as the ‘clean spark spread’. The
dark spread adjusted for the cost of emitting CO2 is referred to as the ‘clean dark spread’.
The CO2 adjustment is the number of metric tonnes of carbon dioxide emitted into
the atmosphere to produce one MWh of electricity. A figure of 0.42 would be typical for
natural gas while the figure applied to coal is 0.85.
The following is a full worked example of how the numbers are derived. Let us
assume that the cost of day-ahead electricity in the UK is quoted at GBP 34.50/MWh and
that day-ahead gas for delivery at the UK National Balancing Point is 36p per therm. The
first step is to divide the price of gas by 100 to express it in GBP/therm; the value is therefore GBP 0.36. To convert therms into MWh, the market uses a value of 0.0293071/therm
so dividing by this value returns a figure of GBP 12.28/MWh. This is the cost of the input
fuel in the electricity generation process.
The next step is to take account of the fuel efficiency which involves dividing the
fuel cost in GBP/MWh by 0.49131. This returns a value of GBP 25.00 rounded to two decimal places. The unadjusted or raw spark spread is therefore GBP 34.50 – GBP 25.00 =
GBP 9.50.
Let us assume the cost of emitting one tonne of CO2 was quoted at GBP 10.75.
Since the CO2 adjustment is 0.42 this equates to a cost of GBP 4.52/tonne of electricity
produced. The spark spread adjusted for the cost of emitting carbon would therefore be
GBP 9.50 – GBP 4.52 = GBP 4.98.
A similar exercise can be performed for the dark spread. Here the assumption is
that steam coal is assumed to have a heat content of 12,000 Btu per pound weight, which
equates to 7 MWh per tonne. To obtain the price per MWh for a short tonne of steam
coal its price is divided by seven.
Assuming the same prices for UK electricity as in the previous example and using
a coal price of USD 71.15 per short ton (API 2, 1st Monday Coal Forward; a short ton is
0.907 of a metric tonne), converted at a spot exchange rate of GBP 1.00 = USD 1.90 gives
a sterling price of GBP 37.45/tonne. The price expressed in equivalent units of MWh
would therefore be GBP 5.35. Using a thermal efficiency of 36% gives the following
dark spread:
Raw Dark Spread = GBP 34.50 − (GBP 5.35∕0.36) = GBP 19.64
301
Electricity
The ‘clean dark spread’ is calculated using the same value for CO2 emissions as in
the previous example but assuming a higher CO2 adjustment of 0.85 (recall that coal
generates more carbon dioxide in the production of one MWh of electricity). The cost
of emitting one tonne of CO2 is assumed to be GBP 10.75.
Clean dark spread = GBP 19.64 − (0.85 × GBP 10.75) = GBP 10.50
If the spread is positive, buying gas and selling electricity will be profitable.
However, if the spread were to go negative an electricity producer who has bought gas
forward on a fixed basis and has sufficient flexibility within their generation capacity
may be encouraged to sell their natural gas back to the market.
More importantly it will indicate the increasing impact of the price of carbon on
the price of power. This effect varies between markets depending on the source of
primary fuel in generation. For example, the impact of the price of carbon is greater
in Germany than in the UK given that there are a greater proportion of coal fired
power generators.
The spreads could be used either by financial institutions to express a view on the
differential between the two types of power or could be used by generators to lock in
a particular margin. If an electricity producer were to sell the spread (sell electricity,
buy natural gas) it would allow them to lock into an advantageous physical margin for
future delivery. The exposure could be subsequently unwound by executing a reversing
position to buy back the spread. This would have to be executed using OTC forwards as
electricity futures are notoriously illiquid.
8.4.6
Marginal heat rates
Another popular measure used often in the US markets is the ‘heat rate’. The concept
of thermal efficiency was outlined earlier in this chapter and the heat rate is closely
related to this as well as the spark spread. In shorthand terms the heat rate is derived as
a ratio of the electricity price and the natural gas price for a particular delivery period.
It is calculated as the price of electricity/price of natural gas.
8.5
TRADING ELECTRICITY – AN OVERVIEW
The main participants in the trading of electricity are utility companies, generators,
suppliers, industrial consumers, and non-physical participants (e.g. banks, hedge
funds, brokers, and trading houses). The country that has the largest population with
a wholesale market for the trading of electricity is the USA. However, trading in this
country has fragmented into a variety of different regional markets that have evolved
in different ways due to the influence of state and federal regulation. In Europe,
the largest consumers of electricity in Europe are Germany, France, the UK, Italy,
and Spain.
302
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
A detailed description of each market for electricity would consume too many
pages and most likely become out of date rather quickly. In this section the aim is to
understand in a somewhat stylised manner:
▪ What is meant by trading electricity?
▪ The motives of the main market participants,
▪ How prices are formed in the market.
An overview of the electricity market and the key trading linkages is shown in
Figure 8.2.
Producers
Their key activities include:
▪ Long-term commercial contracts with suppliers.
▪ Sourcing or disposing of physical supplies on a particular exchange (e.g. EPEX spot)
or via a range of other intermediaries (e.g. energy brokers).
▪ Entering OTC risk management contracts with intermediaries such as commercial
banks.
▪ Sourcing exchange traded derivative solutions via exchange members. For
example, exchanges such as EEX, ICE, and the CME Group offer electricity
futures, but these instruments can only be traded via members of the exchange
such as brokers or banks. These futures have traditionally suffered from a lack of
liquidity.
Suppliers
Their key activities include:
▪ Long-term commercial contracts with producers.
▪ Contracts with retail or industrial consumers.
▪ Procurement or disposal of physical supplies on exchange.
▪ Risk management solutions with intermediaries.
Consumers
Although it is unlikely that a retail consumer would seek to hedge their supply of
electricity, it may be possible to take advantage of certain offers such as a fixed price
supply, which a supplier can offer by using financial swaps. Industrial consumers
could enter risk management solutions on the wholesale market rather than via their
normal supplier.
Electricity
303
Intermediaries
The roles of intermediaries have been touched on, but one additional motivation would
be to use financially settled contracts to express views on expected market movements.
These views could include:
▪ Expressing views on the direction of the prices.
▪ Exploiting price differences between markets.
▪ Making money as a market maker and so profiting from bid-offer spreads.
▪ Providing a brokerage service to generate fees.
▪ Buying and selling structured products to allow investor participants to benefit from
price movements without the need to own the underlying assets.
8.5.1
Load shapes
For contracts that will cover a day or more, how much power is going to be delivered
is clearly specified. This is sometimes referred to as the’load shape’. There are several
standard load shapes, which include:
▪ Baseload – this type of contract provides for the delivery of a constant (i.e. 24 hour)
volume of power for each period (e.g. day, weekend) that the contract covers.
▪ Peak load – although there is no overarching definition of peak load, these contracts
would require the delivery of power between, say, 7:00 a.m.–7:00 p.m., Monday to
Friday when demand is likely to be high.
▪ Off peak – this covers all other periods outside of the peak load definition.
▪ Overnights – these cover periods such as 11:00 p.m.–7:00 a.m.
8.5.2
Contract volumes
When power is traded it may be quoted in MW, which is sometimes referred to as
the contract capacity, contract unit, or unit of power. A Megawatt is an instantaneous
quantity of electricity. If I were to switch on a light bulb that consumes 100 watts, it is
consuming 100 watts of power at any given time that it is on. An energy-efficient light
bulb requires about five watts, a kettle 2,000 watts, and an aluminium smelter a billion
watts (i.e. one gigawatt or a 1,000 megawatts) – about the same energy that is produced
by a medium-sized power plant. A megawatt-hour (MWh) is the volume of power produced or consumed during a particular hour. If the light bulb is switched on for an hour
it will consume 100 watt-hours or 100/1,000,000 MWh. A MWh can be thought of as the
rate of delivery power. A typical household may use in the region of 10 kilowatt-hours
per day.
In trade confirmations the volume per settlement period (e.g. every half hour in the
UK) and the total contracted volume (sum of every half hour) will be stated in MWh
not MW.
304
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
A couple of examples may help clarify. The following is based on the UK market
where each trading day comprises of 48 periods each lasting 30 minutes.
Trade 1: ‘winter peak load 20 MW’ should be interpreted as:
Duration: each half hour period from 7:00 a.m.–7:00 p.m. on Monday to Friday,
from the first day of October to the last day of March inclusive. This is a total of 24,
half-hour periods per weekday.
Total contracted volume: this is calculated as 20 MW × 12 hours × 5 days × 26
weeks = 31, 200 MWh.
Daily volume: this is calculated as 20 MW × 12 hours = 240 MWh
Volume per settlement period: this is the amount to be delivered in each half hour
period of generation and is equal to 10 MWh.
Trade 2: “summer base load 100 MW” is interpreted as:
Duration: each half hour from 11:00 p.m. on the last day of March to 11:00 p.m. on
the last day of September. The contract delivers for 24 hours a day or 48 half hour
generation periods.
Total contracted volume: this is calculated as 100 MW × 24 hours × 7 days × 26
weeks = 436,800 MWh.
Daily volume: this is calculated as 100 MW × 24 hours = 2,400 MWh during
each day.
Volume per settlement period: this is the amount of electricity to be delivered in each
of the 48 half hour blocks and is equal to 50 MWh.
8.5.3
Contract prices and valuations
Although transactions are entered into with the quantity expressed in MW, prices
are agreed in the domestic currency per MWh. When valuing a contract, the total
contracted volume (MWh) is multiplied by the price to find the value of the contract.
Take for example, trade 1 outlined previously. If the contract had been traded at GBP
14.00/MWh, the total value of the contract would be expressed as GBP 14.00/MWh ×
31,200 MWh (total contracted volume) = GBP 436,800. In trade 2 if the contract price
had been negotiated at GBP 15.00/MWh the total contract value would be calculated
as GBP 15.00/MWh × 436,800 MWh = GBP 6,552,000.
8.5.4
Price formation
The main way in which electricity was traditionally priced in a regulated market was
through some mechanism that allowed for operational costs to be covered as well as
providing investors with an acceptable return. In Section 8.3 it was argued that there
are two main market models – pools and bilateral contracts, and so the next step is to
consider price formation within these structures; how much should a generator charge?
Irrespective of the market structure, electricity prices are closely related to production costs, that is, the short-term variable costs incurred to generate power. Indeed,
increased competition will often drive the wholesale price down to the short-run
marginal cost level; logically a generator will only sell their production if they can
recover their variable costs. These variable costs are primarily made up of the input
fuel as well as operating and maintenance costs (e.g. labour costs). However different
generation facilities will have different variable cost structures. Renewable sources
305
Electricity
of fuel incur no procurement cost. You do not have to pay to make the sun shine
or the wind blow. So, for this type of plant, their variable costs are operational, and
maintenance based. Nuclear plants may have high fixed costs since they are expensive
to build but low variable costs. Natural gas plants may have higher fuel costs and lower
fixed costs.
To illustrate the principles of price formation, consider the following example,
which is loosely based on a bilateral market structure. Suppose a generator is
approached by a consumer to supply 500 MW for a given period the following day. To
what price should the producer agree? To keep the illustration straightforward we will
assume that this is the only supply contract on the producer’s books. The producer
will need to know what generation capacity is expected to be available for that period
and the cost to produce the requisite amount of energy. The producer estimates that
they will have 1,000 MW of production capability available, split between a variety of
different source fuels shown in Table 8.2:
TABLE 8.2 Hypothetical capacity and operating costs for an electricity producer.
Input fuel
Available capacity (MW)
Operating costs (USD/MWh)
Diesel
Coal
Natural gas
Renewable
Nuclear
40
130
350
120
360
150.00
120.00
110.00
80.00
70.00
The concept of ordering the input fuels according to cost is sometimes referred to
as the ‘merit order’ or ‘supply stack’ and is based on the premise that to optimise their
output, a producer will start with the cheapest source of fuel, which in this example
would be nuclear. Figure 8.6 is a diagrammatic representation of the producer’s
merit order.
Operating costs
($ / MWh)
Diesel
Natural
gas
Coal
Market price
Nuclear
Renewable
Demand
FIGURE 8.6 The merit order and price formation.
Available capacity (MW)
306
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
For ease of illustration it is assumed that demand is constant irrespective of the
price level (‘inelastic demand’). Some readers may reasonably argue that renewable
operating costs should be the lowest in the merit order. However, these can vary considerably depending on the type of renewable energy.
Since the consumer requires 500 MW of power, their demand will be met from three
sources of fuel: nuclear (360 MW), renewable (120 MW) and gas (20 MW). The single
price paid by the consumer will be based on the cost of producing the ‘marginal unit’ of
power, which is the cost of natural gas. Hence the consumer will pay USD 110.00/MWh.
Note that the profit margin for the nuclear and renewable sources could be substantial, but it is important to bear in mind that the cost of building this infrastructure
would have been significant and is not shown in this example. Once these fixed costs are
factored into the calculation, they may not be more profitable than the other generation
assets.
Another shortcoming of the previous example is that it assumes that the price of
the input fuel will remain constant as a function of demand. This is perhaps unlikely
and so a more realistic representation of the merit order shown in Figure 8.7.
From this it is possible to develop a sense of how prices can change:
▪ Demand may change.
▪ Input fuel prices can increase or decrease.
▪ Plant availability may suddenly change.
▪ Lower electricity prices in other regions may lead to an increase in imports.
▪ Higher electricity prices in other locations may make exports more attractive.
Based on this single order the producer has a capacity margin of 500 MW – the difference between what he expects to produce and what he can produce. The amount
of spare capacity can also sometimes be a good indicator of the level of margins that
producers can charge, i.e. the amount by which the price of electricity exceeds their production costs. This margin is sometimes referred to as a scarcity premium and could last
for a single generation period or longer. Levels of reserves or spare capacity are defined
Operating costs
Diesel
Natural
gas
‘The Devil’s
elbow’
Coal
Market price
Renewable
Nuclear
Demand
FIGURE 8.7 The merit order and price formation.
Production capacity
Electricity
307
as (Installed capacity peak demand)/Installed capacity. Market analysts suggest that
once reserve levels fall to about 10–13%, it is likely that there will be strong upward
price pressure with increasing margins (Barclays, 2008).
8.5.5
Optimising production
The chapter on crude oil illustrated how a refiner could optimise their production.
In this section we will illustrate similar principles for the electricity market. The
example is simplified and stylised for clarity purposes. Suppose an electricity producer
decides to forecast how much they expect to supply in some future period; perhaps a
single day in one week’s time. They will be able to build a profile of their forecasted
production based on several factors:
▪ Existing contracts with suppliers,
▪ Forecasted demand,
▪ Historical patterns,
▪ Weather forecasts,
▪ Significant events (e.g. sporting occasions).
The producer will then consider how to meet this production requirement:
▪ Produce the electricity from their own resources.
▪ Buy electricity from another producer. For example, if the market price for supplying electricity is below the producer’s marginal cost it may be more efficient to buy
the power from another producer.
▪ Buy the electricity via an intermediary (e.g. broker or trading house) or via an
exchange in the form of a physically settled forward or future. This will allow them
to procure the supplies that they need as well as immunizing them against price
volatility.
They will be aware that things could change significantly; consumers may choose
to use more or less than expected and so suppliers may have to change their demand
requirements. As they approach the specific day, they will gradually buy and sell
to match their evolving demand pattern. Some market structures have organised
exchanges that offer a physically settled ‘day ahead’ and ‘intraday’ facility to accommodate this type of trading. For example, the EPEX spot exchange in the UK offers
a ‘double-sided blind auction’. Participants who are looking to either buy or sell to
manage their very short-term demand and supply profiles can enter anonymous orders
onto the exchange. The exchange will then run an algorithm that creates binding
contracts based on matched orders.
These principles could be expanded into a longer-term horizon. For example, the
demand for a 12-month period could be managed by:
▪ Buying an annual physically-settled forward contract to cover the minimum
amount that is expected to be generated.
▪ Seasonal or quarterly contracts could be used to procure supplies in the colder part
of the year (e.g. winter and spring).
308
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Monthly contracts could then be used to manage periods that may be exceptionally
cold (e.g. January).
▪ More specific contracts could then be introduced such as off- and on-peak contracts
or weekend baseload.
The use of forwards and futures can also lead to the amount of traded power
exceeding the actual amount that is physically produced and consumed. This could
arise when a participant enters a forward commitment and then subsequently decides
to terminate the transaction prior to settlement. If traded on an exchange, it would be
possible to take an offsetting position, as only net amounts would require delivery. For
example, in 2018, the UK generated 333 TWh, while estimates from the London Energy
Brokers Association suggested that for the same period approximately 884 TWh was
traded on an OTC bilateral basis. Figures for Germany, the most actively traded European market, suggest a physical generation balance of 540 TWh and OTC derivatives
worth 5,330 TWh on a notional basis.
8.5.6
System imbalances
Another important aspect to consider within the trading context is the TSO/ISO.
Suppliers will make a judgment as to how much electricity they require for their end
consumers and then enter contracts with generators to fulfill these commitments.
These contracts could cover short- and long-term requirements. Contracts to buy
and sell electricity are made bilaterally through commercial over-the-counter (OTC)
negotiations or via recognised exchanges. This type of trading will continue until ‘gate
closure’, which will be a defined period (e.g. one hour) before the specific generation
period. Once a contract has been made, the two parties are required to notify the system
operator of the terms of the deal but not necessarily the price. This will allow the
system operator to identify and attribute any imbalances after the specified generation
period. An imbalance occurs when there is a difference between their actual physical
outcome with the trades that they had contractually agreed.
Prior to gate closure generators and suppliers must deliver a final physical notification to the system operator, which represents an expectation of how much they will
supply and consume. Depending on the structure of the market the generators and consumers may additionally indicate how much they are willing to deviate their production
from this indicated level. This is done by submitting bids and offers, which represent
the amount and price that they are willing to vary production from their final physical
notification.
During the period after gate closure as well as during the actual generation period,
the transmission system operator will aim to resolve any imbalances between production and consumption on a real time basis. The imbalance can be managed by utilising
the extra flexibility to consume or produce by the market participants from their notified levels. The system operator may also be faced with spikes in demand, power station
failure, or perhaps physical capacity problems of the transmission grid.
During the half hour generation period, power will be delivered or consumed as
per the agreed underlying consumer contracts agreed by the producers and consumers.
309
Electricity
After this period, any imbalance is calculated by reconciling what was supposed to be
generated or consumed with what was actually consumed. Any surplus or deficit is
charged back to the guilty party by the imbalance system based on the prices the system
operator were faced with when trying to resolve the imbalance.
Imbalances occur for a variety of reasons:
▪ Traders of electricity may buy more or less energy than they have sold.
▪ Generators may produce more or less energy than they have sold.
▪ The customers of suppliers may consume more or less energy than the supplier has
purchased on their behalf.
▪ Failure of generation equipment.
▪ Scheduling errors.
▪ Output variances due to ambient temperatures. This is because energy can be ‘lost’
during transmission due to factors such as temperature. As a rule of thumb these
losses are around 10%. So, if demand is 100 MW, then it is likely that 110 MW will
need to be generated to ensure that the system is in balance.
8.5.7
Timing mismatches
Depending on the structure of the market, there may also be a mismatch in the structural timing of consumer purchases and producer sales. Large power producers may
seek to sell their production on a forward basis for cash flow certainty, perhaps to service a loan facility used to construct the plant. Equally, they may be happy to sell on an
opportunistic basis if they believe that a certain price represents value. Buyers of power
may prefer to buy their power within a shorter forward time frame. This mismatch can
be accommodated by non-physical intermediaries who will be willing to manage the
associated risks.
8.5.8
UK trading conventions
Until 2014, trading in the UK market was based exclusively on the EFA (Electricity
Forward Agreement) calendar. This calendar broke down a year into 12 months each
containing four or five weeks, with every week allocated a particular number. At the
highest level, the trading year is broken down into two seasons with summer defined
as EFA April–EFA September and winter as EFA October–EFA March.
The year was also expressed in quarters:
Quarter 1 = EFA January–EFA March
Quarter 2 = EFA April–EFA June
Quarter 3 = EFA July–EFA September
Quarter 4 = EFA October–EFA December
310
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Equally it was possible to break down the EFA trading year into a series of individual EFA months, but it should be noted that the EFA calendar did not always follow
the regular Gregorian calendar. Every EFA week was designated a particular number
and was broken into EFA days, with a day starting at 23:00 on the previous night to
23:00 on the day in question. Each EFA day was divided into six EFA blocks of four
hours each:
Block 1
23:00–03:00
Block 2
03:00–07:00
Block 3
07:00–11:00
Block 4
11:00–15:00
Block 5
15:00 –19:00
Block 6
19:00–23:00
Each of these blocks could be referred to in abbreviated fashion using ‘WD’ to
denote that the block occurs during a weekday or ‘WE’ for a weekend. So WD3 would
relate to a weekday period covering 07:00–11:00, while WE5 would cover a weekend
period extending from 15:00–19:00.
To further complicate matters each day could be further subdivided into several
half hour periods to reflect the actual physical generation patterns. Block 1 of a particular trading day included the 47th and 48th half-hour periods of the previous day
and periods one to six of the calendar day in question. This structure would be the
same for the remaining EFA daily blocks. These half-hour periods are referred to as
settlement periods.
After 2014 the market widened to incorporate trades based on a traditional
Gregorian calendar to come into line with other European markets. However, contracts
based on the EFA calendar traded before 2014, but with an expiry after this date would
still be required to apply EFA principles. But like many markets, once an idea is firmly
embedded, it can prove to be very difficult to shift and so at the time of writing (early
2020), indices referencing ‘trading blocks’ exactly equal to the EFA approach were still
being quoted in the UK market.
8.5.9
US traded markets – an overview
Although oil markets are global, natural gas markets largely continental, electricity
markets in the USA are regional. Although there may be some sharing of resources
by neighboring markets, the US transmission system was not designed to move large
amounts of power across the country or indeed over very large regions. Nonetheless, what sharing of resources that has occurred has led to a relatively small number
of actively traded markets. The general principles of electricity regulation in the
US traditionally meant that utilities were subject to regulation and often had to prove
that they were able to serve customers many years into the future. As a result, they
would sometimes collaborate with other local entities to share power plants and
transmission lines. This concept of joint planning gave rise to ‘reliability organisations’
Electricity
311
that were mentioned in Section 8.3.4. The traded markets that emerged from deregulation essentially formed around these same organisations and locations as well as
sometimes using the same name (e.g. ERCOT). The main wholesale markets are:
More actively traded
▪ California Independent System Operator (CAISO)
▪ Midcontinent Independent System Operator (MISO)
▪ New England (ISO-NE)
▪ New York (NYISO)
▪ PJM (Pennsylvania, New Jersey and Maryland)
▪ Southwest Power Pool (SPP)
▪ Texas (ERCOT)
Less actively traded
▪ Northwest
▪ Southeast
▪ Southwest
When electricity is traded on a wholesale basis, the contracts are physically
settled for delivery at a pre-agreed location. Within each of these regions there will be
several individual trading points (e.g. PJM-West; Southern California SP-15, Northern
California NP-15).
There is no single national price for electricity in the USA. Physical producers and
consumers will buy and sell electricity in the regions of their physical operations. Power
traders (e.g. non-physical intermediaries) may choose to transact in one or more of the
regional markets. For example, in the Texas market, there are four related, but distinct
sub-regions: North, West, South, and Houston. So, although prices in these regions will
be related in terms of physical capacity, restraints have traditionally prevented lower
cost energy flowing to adjacent areas where prices are higher. Congestion occurs when
lower priced generation is available on one part of the grid, but because of physical
capacity constraints, cannot be delivered into another location. As a result, the receiving
area must source its generation from local facilities that will be more expensive and so
different prices will apply in each location for that generation period. This pricing technique is referred to locational marginal pricing, which is the marginal price for energy
at the location where it is delivered or received. When the lowest priced electricity can
reach all the locations, then theoretically the price of power across the grid will be the
same. For example, on the day of writing this, the PJM1 website listed 22 zones2 , each
with a different price for electricity ranging from USD 40.51/MWh to USD 44.21/MWh.
According to Blumsack (2018), ‘the locational marginal price at some particular point
1
To give some context, PJM coordinates the movement of wholesale electicity in all or parts of 13
US States and the District of Columbia.
2
Individual locations within a grid maybe referred to as nodes, while a zone represents a number
of nodes.
312
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Transmission capacity = 80 MWh
A
B
Demand = 50 MWh
Marginal cost = $10/MWh
Demand = 100 MWh
Marginal cost = $15/WMh
FIGURE 8.8 Example of locational marginal pricing.
in the grid measures the marginal cost of delivering an additional unit of electric energy
(i.e. a marginal unit MWh) to that location’. The following example is based on his text.
Suppose there are two linked locations within a grid: node A and node B. A generator
in node A has a marginal cost of USD 10/MWh, while local demand is 50 MWh. The
generator in node B has a marginal cost of USD 15/MWh and local demand is 100 MWh
(Figure 8.8).
We will assume that both generators have the capacity to produce 200 MWh at their
respective marginal costs. If the transmission line had an infinite supply capacity, then
the system operator would dispatch all the required electricity from the generator in
node A at a cost of USD 10.00/MWh. This generator would be able to meet demand
from both nodes (i.e. 150 MWh) and there would be a single system marginal price of
USD 10.00/MWh. Now let us impose a constraint on the transmission capacity between
the two nodes of 80 MWh. In this case the system operator dispatches the generator in
node A for a total of 130 MWh. This comprises the local demand of 50 MWh and an
amount equal to the inter-node transmission capacity of 80 MWh. However, there is
still 20 MWh of unfilled demand in node B and so the system operator dispatches the
local generator in node B to meet this demand. However, the marginal cost for doing
this is USD 15.00/MWh. The customers in node A will pay USD 10.00/MWh; this is their
LMP. The customers in node B will pay USD 15.00/MWh for all the energy consumed;
this is their LMP. This is even though some of it was supplied from node A at a lower
price. The prices paid by the customers in node B are a little unfair and that some form
of weighted average would be more appropriate. Sadly, the LMP method does not work
that way! From the generators’ perspective, they are each paid their respective marginal
costs based on the amount that is generated.
The system operator collects revenues from the customers in each node as follows:
Node A∶ 50 MWh × USD 10.00∕MWh = USD 500
Node B∶ 100 MWh × USD 15.00∕MWh = USD 1,500
Total collected = USD 1,500
The system operator pays the generators as follows:
Node A∶ 130 MWh × USD 10.00 = USD 1,300
Node B∶ 20 MWh × USD 15.00 = USD 300
Total paid = USD 1,300
Electricity
313
Note that the system operator has collected more than they have paid, and this
difference is referred to as ‘congestion revenue’ (sometimes variously referred to as a
cost, charge, or price). How the system operator allocates this congestion revenue is
beyond the scope of the text.
Financial Transmission Rights (FTRs) are a popular instrument that can be
used when prices vary in different locations. FTRs (also sometimes referred to as
‘congestion revenue rights’, ‘transmission congestion contracts’, or ‘transmission congestion rights’) are ‘contracts for difference’ (CFD) auctioned by a system operator and
represent an agreement whereby the system operator will compensate a participant if
an actual congestion cost between two points on the transmission system is greater
than a pre-agreed level. However, it can also be a liability if the cost is lower than the
pre-agreed level. According to Moody’s (2018) they are ‘financial instruments that
allow market participants to offset potential losses or hedge against the congestion
component of locational marginal prices in day-ahead electricity markets. An FTR
obligation contract entitles the holder to be compensated if congestion occurs between
two points on the grid in the same direction as stated in the contract. The contract
holder is charged if congestion occurs in the opposite direction stated in the contract.’
Returning to our LMP example, suppose that the day before, the generator in
node A wishes to sell electricity into node B, and they enter an FTR in the direction
from node A to node B. If the clearing price in node B is higher than node A, then
the FTR will pay out this differential. So in our example, had the generator in zone A
executed such a deal they would have sold their power in both nodes for which they
would have received the lower of the two prices (i.e. USD 10.00). Since the LMP in
zone B was higher at USD 15.00 the FTR would have paid them a USD 5.00 differential
per unit of volume.
Since these are financially settled instruments, they can be used by participants to
express views on expected congestion costs without any requirement to make or take
physical delivery. They can also be structured to act in a similar way to financially settled
options rather than the CFD variant noted above.
8.6
ELECTRICITY DERIVATIVES
8.6.1
Electricity forwards
Unlike other markets that use a forward curve for pricing contracts for delivery at different points in time, the fact that electricity cannot be stored means that the traditional
methods of pricing a forward contract (i.e. spot price plus net carry) will not apply. As a
result, expected levels of supply and demand determine the forward. Additionally, the
forward price may move independently of the spot price, as there is no mechanism for
arbitraging the two markets. So if a forward contract is perceived to be trading ‘rich’ to
some notion of theoretical value, it is not possible to sell the electricity forward, buy it for
spot delivery, store it for a given period, and then use it to fulfill the forward obligation.
Over-the-counter forward deals may be attractive for a variety of reasons:
▪ Price certainty is achieved.
▪ Avoids the need to buy in a volatile spot market.
▪ Physically settled transactions can be a useful tool to secure supplies.
314
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Can be executed on a ‘paper’ rather than physical basis so that the supply and the
risk management can be separated.
▪ Can be used to speculate on movements in the market.
▪ Certain financing transactions may require the end user to use forwards to give
certainty of revenues.
A fixed price power forward is straightforward in terms of how the contract will be
set out. Although each contract will vary the basic terms will be similar. For example,
the following criteria need to be specified for a physically settled transaction:
▪ Trade date
▪ Buyer
▪ Seller
▪ Type of governing contract
▪ Price
▪ Volume
▪ Load shape
▪ Delivery schedule
▪ Delivery point
The contracts will often contain a schedule outlining the detail of the supply. Take
for example the transaction for delivery on the RTE Grid in France shown in Figure 8.3.
The contract fixes the price for delivery of 50 MW for 12, one-hour periods to give
a total contract amount of 600 MWh. At an agreed fixed rate of EUR 60/MWh, the total
value of the contract is set at EUR 36,000.
TABLE 8.3 Summary of physical supply contract.
Total
Supply
Period
Start
Contract Contract
Applicable Days
From To Capacity quantity
End Mon Tue Wed Thu Fri Sat Sun CET CET (MW) (MWh)
3 Aug 3 Aug
X
08:00 20:00
50
Contract
Total
price
(EUR/ amount
MWh) (EUR)
600
60
36,000
Another example of a physical forward contract requiring the delivery of a fixed
volume of power at an agreed price for an agreed period is given below.
Trade date:
Effective date:
Maturity:
Buyer:
Seller:
Price:
Volume per hour:
Total Volume:
Load shape:
Delivery point:
26 June
27 June
28 June
Counterparty A
Counterparty B
GBP 32.00/MWh
50 MWh
1,200 MWh
Baseload
GB National Balancing Point
315
Electricity
A load shape refers to the level of expected demand during certain times of the day.
A baseload contract provides for the delivery of a constant volume of power throughout
the day. A peak load contract for the UK typically runs from 7:00 a.m.–7:00 p.m. on a
weekday. The other popular standard load shape is off peak, which is defined as everything except the peak period. Since this forward is a baseload contract it requires the
seller to deliver a constant volume of power for a 24-hour period. In this example the
contract covers a one-day period in one-day’s time, but since these contracts are traded
on an over-the-counter basis it would be possible for the counterparties to the trade to
agree a longer period if necessary.
Floating price forwards
Floating price forwards, or ‘index price forwards’ allow the parties to the transaction to
deliver a given amount of MW at a given time period in the future at a price that will be
set at the point of delivery based on an agreed market price index.
Cross market trading
One type of trade growing in popularity allows a trader to exploit the difference in power
prices between countries. Interconnectors allow participants to move electricity across
countries, but to do so the trader must have purchased the right to transmit the power.
The capacity to transmit is auctioned on a regular basis by the grid operator. Cross market trades are in essence very simple. Suppose a trader sees that the price of power in
France is EUR 55.00/MWh while the price in Germany is EUR 60.00/MWh and that
the trader has purchased annual transmission capacity at a cost of EUR 3.00/MWh.
He agrees a trade size of 100 MW to buy the power in France and sell it into Germany.
An annual baseload contract would generate a profit of:
EUR 60.00 − (EUR 55.00 + EUR 3.00) × 24 hours × 365 days × 100
MW = EUR 1,752, 000
8.6.2
Electricity swaps
As entities such as hedge funds do not trade physical power, financially settled swaps
represent an alternative way of taking exposure to the underlying market. These types
of swap involve the exchange of cash flows where one cash flow is fixed while the other
is floating. The floating cash flows reference a market index such as the LEBA (London
Energy Brokers Association) day-ahead index. This index price is constructed as the
volume, which is the weighted average of all day-ahead baseload trades executed in
London by several contributing institutions. GB power swaps settle against the average
value of the index through the period being traded, typically months, quarters, seasons,
or years.
Suppose a hedge fund wishes to express a view on anticipated power market movements over the next winter. They decide to ‘buy’ 30 MW of winter baseload power at
GBP 45.50/MWh, referencing to the LEBA day ahead index with monthly settlement
covering the six-month period from October–March.
316
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Payments will take place at the end of the month against the average of the day
ahead index values published every day throughout the month in question. To illustrate
the cash flow amounts let us assume that one of the months has 31 days and that the
floating LEBA index has fixed at GBP 40.00. Since the hedge fund has ‘bought’ the swap
they will pay the fixed price and receive the floating index. The cash flows will be:
Fixed cash flows paid
30 MW × GBP 45.50 × 24 hours × 31 days = GBP 1,015,560
Floating cash flows received
30 MW × GBP 40.00 × 24 hours × 31 days = GBP 892,800
Since the timing of both cash flows coincides, it is conventional to make a net
payment. Hence the hedge fund will make a net payment of GBP 122,760.
So, what would be the hedge fund’s motivation for executing this swap transaction?
In simple terms, the swap price can be thought of as a weighted average of the forward
prices for each of the periods that the swap covers. From the hedge fund’s perspective,
the trade is essentially bullish, as they must believe that the actual day-ahead price will
on average be greater than the individual forward rates used to initially price the swap.
So in the previous example the fixed price would be a function of the forward prices
for each individual month within the winter period. As such, the trader’s motivation is
simple – they will always aim to ‘beat the forward price!’
Swaps can also be used for hedging purposes. For example, a supplier may have
agreed a floating price supply contract with a generator but wishes to transform their
exposure such that on a net basis they would pay a fixed price. In this case, they could
receive a floating cash flow from a swap provider to negate the underlying exposure
with the generator and in return, agree to pay a fixed price under the terms of the
derivative. Equally, based on the same parameters a generator could also transform
their market risk by paying floating and receiving fixed under the terms of a swap.
The generator would receive a variable price for the physical power they sell which
can finance the floating swap payment. Their net revenue is therefore the fixed price
received under the swap.
Continental European power swaps will have several key features with differences
to reflect local market conventions. The main markets that will be traded include
France, Germany, Spain, Scandinavia, and the Netherlands. However, a few generalisations can be made as to their structure. There will be a stated maturity such as six
months as well the frequency with which the cash flows will be paid. This will usually
be monthly and is referred to as the calculation period. The contract will outline the
load shape, which may typically include:
▪ Baseload – 24 one-hour periods from 00:00–24:00.
▪ Peak load – 12 one-hour periods from 08:00–20:00.
▪ Off-peak – 12 one-hour periods from 00:00–08:00 and 20:00–24:00 Monday–Friday;
24 one-hour periods from 00:00–24:00 Saturday and Sunday.
317
Electricity
Payment is made in arrears after each calculation period usually with a lag, which
could be either 5 days or 20 days.
In terms of the ‘price’ of the swap, the fixed rate per MWh will be stated as well
as the counterparty responsible for making the payment. For continental power swaps,
the floating payment may typically be based on hourly prices. The counterparties to the
trade would agree a notional quantity for each hour of generation expressed in MWh,
which is applied to the floating price derived from the agreed source. The floating payment made at the end of each calculation period is the sum of these monetary values.
So as an example, take a 92-day baseload contract with an agreed notional quantity
of 6 MWh for each hour of generation. At the end of each month the counterparties
would have to take the hourly price observations, apply them to this notional quantity,
and add them all up to calculate the floating amount. The total notional quantity for
the contract would be equivalent to 13,248 MWh (6 MWh × 24 hours × 92 days).
8.6.3
Contracts for difference
As a way of encouraging low carbon power generation, the UK government has
mandated a new product referred to as a contract for difference (CFD), illustrated in
Figure 8.9.
The low carbon contracts company (LCCC) is a government-owned entity and acts
as the CFD counterparty to the generator. The generator will produce their electricity
and sell it into the market at the prevailing market price. To help them manage market price movements they could then enter a CFD for a period of 15 years. Under the
terms of the CFD the generator will pay or receive a cash flow based on the difference
between a pre-agreed strike price and a variable ‘reference price’. The strike price is
set by the government and will vary according to the type of technology. For the years
2015/2016 the strike price for onshore wind technology (OWT) was set at a maximum of
GBP 95.00/MWh. However, there are circumstances where CFDs are granted on a competitive basis, so this strike price acts as a maximum value. For the period 2015/2016 the
fixed rate payable by the LCCC for OWT contracts was set at GBP 79.23/MWh. The variable reference price payable by the generator is based on the average market price for
electricity.
Suppose an onshore wind generator sells electricity in the open market and receives
GBP 70.00/MWh. Under the terms of the CFD and assuming a fixed strike of GBP
79.23/MWh they will receive a supplemental payment of GBP 9.23/MWh such that their
Sale of electricity
Strike price
Low carbon
contracts
company
Low carbon
generator
Market price
FIGURE 8.9 Example of contract for difference.
Reference price
318
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
income will be equal to the CFD’s strike price. If the price the generator received for the
sale of their electricity was, say, GBP 85.00/MWh, then under the terms of the CFD they
would be required to pay GBP 5.77/MWh (i.e. GBP 79.23 – GBP 85.00). However, in both
scenarios the generator’s net income from both sides of the transaction would be equal
to the CFD’s strike price and so they are protected from volatile wholesale prices.
8.6.4
Swaptions
Consumer hedge – manufacturing company
A swaption is defined as an option on a swap. Consider the following hypothetical
example. A large UK manufacturer is looking to lock in the purchase price of its physical supply of power. Looking at forward prices, they believe the actual price of power
will be, on average, higher than the forward prices currently being quoted in the market.
They currently consume about 25 million kWh per month of power and are looking to
hedge half of this volume. They enter a fixed price swap with the following terms:
Swap buyer:
Swap seller:
Volume:
Tenor:
Fixed price:
Reference price:
Settlement frequency:
Manufacturing company
Bank
12,500,000 kWh per month
12 months
GBP 50.00/MWh
GB baseload
Monthly
On the last business day of each month the parties would calculate the reference price which, as was illustrated in the last section, would usually be an average of
daily prices observed over the month.
The swap would be financially settled so as not to interfere with the manufacturer’s
physical supply relationship. To illustrate the impact of the swap let us consider some
possible outcomes. Let us assume that in one month the average cost of GB baseload
power was GBP 55.00/MWh. The manufacturer would pay this amount to his physical
supplier and receive the agreed amount of power. Under the terms of the swap the bank
would be required to pay this same amount to the manufacturer and in return receive
the pre-agreed fixed price of GBP 50.00/MWh. Since both cash flows under the swap
coincide there would be a net settlement of GBP 5.00/MWh in favour of the manufacturing company. This means the net cost of buying their power is GBP 50.00/MWh (GBP
55.00/MWh paid out to the supplier with GBP 5.00/MWh received under the swap).
If the average price were, say, GBP 40.00/MWh then the manufacturer would pay this
amount to their supplier and be required to make a net payment of GBP 10.00/MWh
under the terms of the swap. The net result in both examples is that the client achieves
a fixed price of GBP 50.00/MWh – the same value as the fixed price in the swap.
Had the trade been structured as a swaption, the manufacturer could purchase the
right to ‘buy’ a swap (i.e. purchase the right to pay a fixed price3 ). Like the options
3
This could also be referred to as the purchase of a payer swaption.
Electricity
319
considered earlier, this transaction would also incur a premium. At first glance it may
appear that this transaction would be appropriate if the client believed that power prices
were going to rise as they could exercise the option, enter into the swap, and then pay
a fixed price. However, this is not the case. If they were certain that prices were going
to rise, they would simply enter the swap; entering a swaption will incur a premium
whereas the swap is traded on a bid-offer spread. Arguably, the right motivation for the
transaction is that the client believes that power prices are going to fall, but decide to
take out protection in case their view is incorrect.
It is also worth noting that few corporates are willing buyers of options.
Anecdotally, their main objection to single option structures is the requirement to
pay a premium. As a result, it is most common for banks to structure zero premium
transactions where different options are combined such that the client does not incur
any upfront charge. A very popular zero premium transaction is termed a ‘min-max’
structure (sometimes referred to as a zero-premium collar). The purchased swaption
we considered earlier fixes the maximum price that the client will pay if they decide
to exercise the option. To achieve a zero premium, the client not only buys this option
but also sells a second option to the bank, which if exercised, would require the client
to pay a fixed price albeit at a lower strike. The strike of the sold option is set at a
level such that the premium generated fully finances the cost of the purchased option.
The sold option fixes the minimum price that the client will pay and would only be
exercised by the bank against the client if the price of power is falling. This will prevent
the client from enjoying the benefits of lower prices although it should be remembered
that they have protection from prices rising at no premium cost. If the price of power
is between the upper and the lower strike prices, neither option is exercised, and the
client buys their power at the prevailing market price. Given many clients’ wariness of
selling options, the transaction term sheets rarely mention the fact that the structure
involves the sale of an option and may simply talk about a ‘cap’ and a ‘floor’ price.
8.6.5
Spread options
The mechanics of spread options were considered in Chapter 1. How could they be
applied within the context of electricity?
In Section 8.4.5., the concept of spark and dark spreads were introduced. A possible ‘view-driven’ strategy would be to execute an option that pays off if the underlying
spread were to increase or decrease relative to a pre-agreed strike price.
Another application is one where a client could be protected from an adverse movement in the price of power in another country. Suppose that a UK chemical producer
has a competitor based in France. In this scenario, power is a significant input to the
process and their competitive position could harmed if UK prices rise relative to those
in France. A spread option that pays off if the price of French power declines relative to
those in the UK could be a solution.
8.6.6
Monetising embedded optionality
Electricity represents a significant cost for an aluminium producer. It is possible that
in a deregulated electricity market where the aluminium producer is buying its physical power on a floating price basis, that a spike in power prices could force a potential
320
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
shutdown. If the producer has a flexible power supply contract that grants them the
ability to sell on their electricity, they have a form of embedded optionality that could
be monetised (Barclays, 2003).
When considering the cost of closing a plant, an aluminium producer might
consider:
▪ The requirement to keep a certain number of people employed to maintain the
plant.
▪ The profit or loss that might be generated from the resale of their raw material,
alumina.
▪ The operational costs of physically shutting down and restarting the plant.
▪ The cost of breaching any existing contractual supply contracts.
Once these costs have been established it would be possible to calculate an ‘indifference curve’. This would represent a series of aluminium and power price combinations
where the producer would be indifferent to whether they will:
▪ Purchase power and produce aluminium or,
▪ Close down aluminium production and on sell their electricity.
This relationship is shown in Figure 8.10.
This form of optionality can be monetized by the sale of a call option. That is, the
aluminium producer sells the right that would allow the option buyer to buy electricity
at a pre-agreed strike price and a pre-agreed time in the future. The option is structured
in such a way that it will comprise of a time-limited option (e.g. three months), which
if exercised, would then result in a forward as opposed to a spot sale of electricity. This
delayed sale of electricity would give the aluminium producer sufficient time to shut
down their operations.
To illustrate how this would work consider the following simplified example.
With the current price of electricity at USD 80.00/MWh, an aluminium producer sells
Price of
electricity
Sell electricity
‘Indifference
curve’
Sell aluminium
Price of
aluminium
FIGURE 8.10 Aluminium producer’s indifference curve.
Electricity
321
a three-month out-of-the-money call option that references electricity. If exercised,
it would allow the option buyer to receive a certain volume of electricity at a pre-agreed
price, say, USD 100.00/MWh. It is assumed that this strike was set at or above their
indifference curve. The option would be struck OTM, as it will be exercised only if
the price of electricity increases significantly, and so as a result the premium would
be relatively low. In this example let us suppose that the option premium is USD
3.00/MWh. The following scenarios would then result:
Price of electricity is below the strike price
Buyer does not exercise the option. Option expires worthless and aluminium producer
retains premium, which can be used to enhance the company’s cash flow. Producer opts
to produce aluminium.
Price of electricity is equal to the strike price
Buyer exercises the option and takes delivery of the agreed volume of electricity at the
strike price of USD 100.00. Aluminium producer pays USD 100.00 for the electricity
from its supplier but retains the premium of USD 3.00/MWh.
Price of electricity is above the strike
Buyer exercises and takes delivery of the electricity as before. However, if the price of
electricity has exceeded the retained USD 3.00/MWh premium, then the aluminium
producer must now pay more to their electricity supplier than they will receive from the
option buyer. If they have hedged 100% of their power supply this could incur losses.
They may therefore decide to execute the option for part of their electricity consumption. The non-option element of any electricity could then be sold on at a higher price
in the open market.
One point that comes out of this is that the aluminium producer would find the sale
of optionality particularly attractive under conditions of higher implied volatility. The
higher the implied volatility of the underlying electricity price, the higher the premium
they would receive.
The example illustrates that like most strategies, the selling of optionality is not
completely risk free. One aspect not considered is if the indifference curve moved after
the option transaction date, but before the expiry of the option. This could be because
of a rise in aluminium prices, which for a given cost of power would push the producer
towards selling the metal instead. One possible solution to this problem would be to create aluminium options that are contingent on the electricity price. One example of this
type of transaction was considered is the two-asset barrier option, which was discussed
in the chapters on gold and crude oil.
8.6.7
Ratio swap on power and aluminium
Power is a key input in the aluminium process and is arguably the most significant
cost faced by a smelter. Conventionally, a smelter may decide to hedge the sale of their
322
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Sale of physical aluminium
Floating price AL
Physical Aluminium
Floating price AL
Fixed price AL
Aluminium
smelter
Fixed price PWR
Bank
Floating price PWR
Floating price PWR
Physical Power
Purchase of physical energy
FIGURE 8.11 Ratio swap.
production and the power exposures separately. However, there is a risk that if the
two hedges are not constructed to complement each other, then the smelter will suffer reduced revenues if one side of the transaction moves against them. One possible
way to overcome this is by means of a ratio swap, which combines both elements. This
is illustrated in Figure 8.11.
The aluminium smelter sells their physical metal and receives the prevailing market price. This floating payment is then swapped for a fixed price under the terms of the
swap. At the same time, they buy the power they require in proportion to the amount
of metal that they wish to sell (e.g. 14 MWh of power for every 1 MT of aluminium that
is sold). Again, this is paid for on a floating basis, but this expense is transformed into a
fixed cost under the terms of the swap. In a perfect world the net impact of the swap is
that all the floating payments cancel out and the smelter is left selling a given volume
of metal and buying the required power at two fixed prices.
8.6.8
Monthly and daily power swaps
In the US market ‘power swaps’ are agreements to exchange a fixed price for the delivery of physical electricity in a particular region for a pre-agreed period. The use of the
term ‘swap’ is potentially misleading as they are arguably more akin to a physically
settled forward.
323
Electricity
A monthly power swap will fix the price for physical delivery in a particular region
for a particular month. If a market participant were to execute a ‘PJM monthly July
swap at USD 50.00/MWh’ then the buyer (i.e. the payer of the fixed rate) would pay
USD 50.00 for every MWh of an agreed volume of electricity delivered in this location
during the specified month. The agreement will typically settle just before the start of
the applicable delivery month.
A daily power swap works in a similar fashion, but, as the name suggests, it will
cover the delivery of power in a particular region for a specified single day (e.g. 1 July).
This transaction would settle the day before the scheduled delivery.
8.6.9
Options on power swaps
This type of option may also be referred to as a ‘power swaption’ and could be structured
with either monthly or daily settlement. The underlying in this case would be either a
monthly or daily power swap.
Monthly swaptions
A call on a monthly power swap would give the buyer the right to receive an agreed
volume of electricity at a pre-agreed fixed strike price for a one-month period.
Expressing this in conventional option payoff terms would be:
Value of call at maturity = MAX (Underlying price − strike, 0)
In this case the underlying price would be the prevailing price for a power monthly
swap that covers the referenced month. Suppose that in the middle of a particular year
a market participant buys the following option:
Option maturity:
Option type:
Option style:
Underlying:
Market:
Load shape:
Delivery period:
Volume:
Total volume:
Strike:
Option premium:
Settlement:
One-month option
Call
European
Monthly power swap
PJM Western Hub (Pennsylvania, New Jersey, and Maryland)
Peak hours (07:00–23:00 - 16 hours, Monday–Friday)
1–31 July (excluding weekends, 23 working days)
100 MWh
36,800 MWh (100 MW × 16 hours × 23 days)
USD 50.00/MWh (ATM)
USD 5.00/MWh
Financial settlement against CME Group ‘PJM Western hub,
Peak calendar month future’
The key features are that it is a one-month option exercisable into a one-month
delivery period that financially settles against an observable futures price. Note that the
maturity of the option is relatively short. One of the reasons for this is that options that
324
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
reference electricity will tend to trade at very high implied volatilities; values of over
100% are not uncommon.
A monthly put swaption would work in a similar fashion although the expiry payoff
would be:
MAX (Strike − underlying price, 0)
Again, it may well be financially settled, but economically the put buyer would
‘deliver’ power in exchange for receiving the strike price.
Daily swaptions
A daily swaption would have similar characteristics to the monthly structure. The main
difference would be that the underlying would be a daily power swap.
These transactions comprise of a series of daily options covering, say, a one-month
period. In the monthly swaption structure there was a single option that covered
23 working days of a calendar month. An option structure that gives the holder the
right of exercise on every one of those days means there are 23 individual options
packaged as a single transaction.
For readers who are familiar with financial markets, this is like the interest
rate cap (i.e. call) and floor (i.e. put) markets. A cap (floor) that covers a predefined
number of periods is made up of a strip of individual options called caplets (floorlets).
Readers who require more information on this product are referred to Schofield and
Bowler (2011).
The economics of option pricing means that a single option based on our example
of 23 days (a monthly swaption) has a lower premium than 23 options, each covering
just one day (a daily swaption for an entire month). Intuitively with the single option
structure there can only be one opportunity to pay out; the multi-option structure can
pay out up to 23 times and will therefore cost more.
Applications
An electricity generator could sell put swaptions to boost their cash flows, albeit with
some degree of risk. Diagrammatically, the economics of the transaction are shown in
Figure 8.12.
It is more of an aggressive transaction as can be seen from the associated cash flows,
which are shown in Table 8.4. The option strike is USD 50.00/MWh, and the premium
is assumed to be USD 5.00/MWh.
If the market price for electricity stays above the strike price, the generator retains
the premium and is not required to make a payment under the terms of the option. The
premium then can be used to improve cash flow. However, once the option becomes
in-the-money note that the net cash flow receivable by the generator falls rapidly. As
the price of electricity falls, not only would they receive lower income from their sale of
physical power, but also they are required to pay out under the terms of the option.
One possible hedging application would be for a wholesale power distributor to buy
call swaptions that would yield protection against a spike in wholesale prices that may
not easily be passed onto their end clients. The call gives the holder the right to ‘buy’
325
Electricity
Sale of put swaption
Sale of physical energy
Strike price
Market price
Electricity
consumer
Generator
MWh
MWh
Premium
FIGURE 8.12 Sale of put swaption by electricity generator.
TABLE 8.4 Payoff profile for an electricity producer who has sold a put swaption.
Income from
selling physical
power (USD/MWh)
+60
+55
+50
+45
+40
+35
+30
Premium received
from sale of option
(USD/MWh)
Option payout
(USD)
Net cash flow
received (USD)
+5
+5
+5
+5
+5
+5
+5
0
0
0
−5
−10
−15
−20
+65
+60
+55
+45
+35
+25
+15
the underlying swap, i.e. pay fixed & receive floating. If prices were to increase sharply
leading to an increase in the cost of their physical purchases, the floating leg of the swap
would finance these higher payments. The distributor’s net cost of buying power would
therefore be the swap’s fixed price.
From a ‘view driven’ perspective, market participants could trade options to express
views on the volatility of power as well as the volatility of the factors that determine
power prices. These might include the prices of input fuels such as natural gas, coal, and
weather. If they were looking to neutralise themselves against directional movements
in the underlying price, the appropriate delta hedge would be a forward-starting swap.
The remaining Greeks such as vega, gamma, and theta could be hedged using power
options with alternative maturities or other delivery locations. Some traders choose to
use options on natural gas to hedge these exposures as they are considered to have better
liquidity. However, this introduces an element of basis risk as the trader would now be
reliant on the correlation between gas and electricity prices remaining stable.
Another interesting point is to appreciate that power swap prices may behave very
differently depending on the region that they reference. For a region like PJM, which
uses a wide variety of fuel types the swap price may vary considerably. Those regions
that rely on a single fuel input such as natural gas may not experience the same degree
of price volatility.
326
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
8.6.10
Heat rate derivatives
Defining the heat rate
Section 8.1.2 introduced the concept of thermal efficiency. This was a measure that
related the electrical energy produced to the energy content of the input. So if 100
units of natural gas produced 47 units of electricity then the thermal efficiency was
47%. Stated as a formula:
Electricity produced∕Energy content of the input
A related measure is the heat rate, which measures the number of MMBTu needed
for a marginal generating plant to generate 1 MWh of electricity4 . In general terms it is
expressed as:
Total input fuel used∕total energy produced
Where natural gas is the input fuel the expression would be:
Input energy (MMBtu∕hour)∕Output power (MW)
So, the heat rate of a power plant expresses how much input fuel is necessary (measured in millions of British thermal unit) to produce one unit of energy (measured in
megawatt hours).
The heat rate is simply the inverse of thermal efficiency and so it follows that a lower
heat rate indicates a higher thermal efficiency and therefore a more efficient generation
system.
Suppose that the energy input was 3,500,000,000 BTUs/hour and this produced
500,000 kW of energy. This would yield a heat rate of 7,000 BTU/kWh. This could also
be expressed as 3,500 MMBtu and 500 MW to give a heat rate of 7 MMBtu/MWh.
To illustrate the concept, consider the relationship between thermal efficiency
and heat rates using the values suggested earlier. These examples are based on the
convention that 1 kWh of electricity has an equivalent BTU content of 3,412.
▪ A thermal efficiency of 100% implies a heat rate of 3,412 BTU/kWh.
▪ A thermal efficiency of 47% (e.g. natural gas) implies a heat rate of 7,2660 BTU/kWh
(47% = 3,412 / 7,260).
▪ A thermal efficiency of 35% (e.g. coal) implies a heat rate of 9,477 BTU/kWh (36%
= 3,412 / 9,477).
In the electricity trading markets heat rates are expressed in price terms:
Price of electricity (USD/MWh) / Price of the generating fuel (e.g. natural gas in
USD/MMBtu)
However, the currency component cancels out and the result is again expressed as
MMBtu/MWh.
4
The measure can also be reported as Btu/kWh.
327
Electricity
Kaminski (2012) makes a useful distinction between the two measures. The ratio
of the two market prices is termed the ‘market heat rate’ (also sometimes called the
‘implied heat rate’), while the ratio of the energy input to the power output is the ‘technological heat rate’. He argues that the technological heat rate of the marginal generator
will determine the marginal cost of electricity in a particular market. For example, if
coal is used instead of natural gas the technological heat rate will change, and since
they trade at different prices, this may have an impact on the price charged for electricity. In a perfectly competitive market, the cost of power will be determined by the
marginal cost of production so the technological heat rate would be equal to the market
heat rate. He points out that the two measures will often diverge:
▪ Marginal cost will typically include an element for operational and maintenance
charges, which would be ignored by the financial markets.
▪ If the generator has some degree of market power, they may be able to set prices
above marginal costs.
Spark spreads were introduced in Section 8.4.5 and in a simplified form they were
calculated as:
Price of power − (price of input fuel∕thermal efficiency)
Given the relationship between thermal efficiency and the heat rate this relationship could also be expressed as:
Price of power − (price of input fuel × heat rate)
Price reporting agencies will often report market heat rates and to illustrate how
these figures would look consider the following example based on day-ahead prices for
electricity and natural gas. The example uses S&P Platts, North American electricity,
methodology and specifications guide.
▪ Price of electricity = USD 39.75/MWh
▪ Price of natural gas = USD 2.89/MMBtu
▪ Marginal (e.g. market) heat rate = 13.75 MMBtu/MWh. This might also be reported
as 13,754 Btu/kWh
They also report spark spreads in USD/MWh:
Electricity price (USD/MWh) −[natural gas price (USD/MMBtu)
× heat rate (MMBtu/MWh)]
The reported values use a variety of different heat rates depending on the
regional market being analysed (e.g. 7,000, 10,000, 12,000, 15,000; all expressed as
Btu/kWh).
In our example, using a heat rate of 7 MMBtu/MWh this would give a spark spread
of USD 19.49/MWh. (39.75 − (2.89 × 7))
328
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Heat rate options
Although these instruments could be structured with coal as the underlying input fuel,
the following examples are all based on natural gas. A European-style heat rate call
option written on natural gas has a fixed heat rate strike. At maturity, this option gives
the buyer the right but not the obligation to pay an agreed underlying price for natural
gas multiplied by the fixed heat rate strike and receive the price of one unit of electricity.
The payoff of a call option at maturity is therefore:
MAX [(Underlying price of electricity
− (Heat rate strike × Underlying price of natural gas)), 0]
Comparing this expression to the calculation of the spark spread using heat rates
indicates that this payoff represents a call option on spark spreads.
A European-style heat rate put option written on natural gas has a fixed heat rate
strike. At maturity, this option gives the buyer the right but not the obligation to receive
the underlying price of natural gas multiplied by the fixed heat rate strike and pay the
price of one unit of electricity. The payoff of the put option at maturity is therefore:
MAX [((Heat rate strike × Underlying price of natural gas)
− Underlying price of electricity), 0]
A typical transaction may contain the following terms:
Option buyer:
Option seller:
Option style:
Option type:
Maturity:
Notional Quantity:
Calculation period:
Total volume:
Determination period:
Monthly premium:
Total premium:
Premium payment:
Commodity A:
Client
Bank
European
Daily heat rate call option
One calendar year
100 MW per hour during each Calculation Period, equal to
a total of 1,500 MWh for each calculation period.
Daily (excluding weekends and holidays) covering the peak
15 hours of 8:00 to 23:00, Eastern Time. (Number of
calculation periods in this example is taken to be 250 for
ease of illustration).
375,000 MWh during the option term (1,500 MWh per
calculation period × 250 days)
Every month during the term of the option
USD 11,000
USD 132,000
The option buyer will make a premium payment to the
option seller, in arrears each month as part of the net
settlement amount.
Electricity
329
Electricity
Floating price A:
For each calculation period (i.e. daily) the arithmetic
average of the day-ahead price for each hour as published
by the ISO New England.
Floating price B:
For each calculation period: (Heat rate × Natural Gas price)
+ variable charge. The natural gas price will be the Henry
Hub Natural Gas in USD per MMBtu. The Heat rate will
be 7 MMBtu/MWh. The variable charge will be fixed at
USD 3.50/MWh and accounts for the start up or
maintenance costs of a plant.
Net Settlement Amount: For each calculation period a settlement amount shall be
calculated as follows:
MAX [(Notional quantity × (Floating price A − Floating price B)), 0]
The settlement amounts for each calculation period (i.e. daily) during a determination period (i.e. monthly) shall be aggregated into a single net amount. In
addition, the option premium payment will be included as a negative amount.
If the figure is positive, the option seller will pay the option buyer. If the figure
is negative the option buyer shall pay the absolute value to the option seller.
The following calculations convey a sense of how the daily settlement amount
would be calculated. The analysis aims to show the impact of relative price changes.
For ease of illustration, the variable charge and the premium have been ignored and
the calculation is based on a single unit of the underlying. We will assume that the
current price of electricity is USD 35.00/MWh and natural gas is USD 5.00/MMBtu.
The ratio of the two prices returns a ‘market heat rate’ of 7 MMBtu/MWh, which is
equal to the strike, so the option is initially at-the-money.
Day #1 – spark spread increases
Average price of day-ahead electricity = USD 40.00∕MWh
Price of natural gas unchanged at USD 5.00∕MMBtu
Implied market heat rate (i.e.ratio of the two prices) = 8 MMBtu∕MWh
Settlement amount = MAX ((40 − (7 × 5)), 0) = USD 5.00
Day #2 – spark spread unchanged from original value
Average price of day-ahead electricity = USD 35.00∕MWh
Price of natural gas unchanged at USD 5.00∕MMBtu
Implied market heat rate (i.e.ratio of the two prices) = 7 MMBtu∕MWh
Settlement amount = MAX ((35 − (7 × 5)), 0)
330
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Day #3 – spark spread decreases from original value
Average price of ‘day-ahead’ electricity = USD 30.00∕MWh
Price of natural gas unchanged at USD 5.00∕MMBtu
Implied market heat rate (i.e. ratio of the two prices) = 6 MMBtu∕MWh
Settlement amount = MAX ((30 − (7 × 5)), 0)
From this analysis it follows that:
▪ The payoffs shown above would have to be multiplied by the notional quantity (i.e.
1,500 MWh for each day).
▪ This type of calculation will be performed on each applicable day (the ‘calculation
period’).
▪ If the variable charge were to be included with the input cost of the natural gas,
then the option payout would be reduced further.
At the end of the month (‘the determination period’), each daily calculation would
be summed and the monthly premium payable by the option buyer subtracted. A positive sum means seller pays buyer; a negative sum means buyer pays seller the absolute
value as the option has expired out-of-the-money and the payment made represents the
premium due.
Hedging applications
A heat rate option is designed to mimic a generation asset. Market implied heat rates
can vary considerably depending on the location. If a particular region has capacity to
use different fuel types (e.g. nuclear, natural gas, coal) then the market heat rate will
change as the source fuel changes. It follows that regions that rely mainly on a single
source of fuel (e.g. natural gas) would have a less variable heat rate.
A generator could buy a put option on the heat rate which would pay off if the
spread between the two fuel prices were to decrease. The fall in their physical operating
margins would be offset by the option payout. Another possibility would see the generator selling OTM call options. If the spread between the two prices were to increase, this
would represent an increase in their physical operating margins. However, once prices
move beyond the strike they would be faced with an increasing payout on the option,
which would negate any additional profit. Typically, the strike rate of the option should
be set at such a level that its exercise would be unlikely. However, the further away the
strike rate from the current value, the lower the premium the generator would earn.
View driven applications
This type of option is an example of a spread option, and so there is a correlation exposure between electricity and whatever input fuel is chosen. The buyer of a call on the
spread will see their option increase in value if the correlation between electricity and
Electricity
331
natural gas were to decrease. This would suggest that the two prices would move in
opposite directions making a payoff more likely.
From a valuation perspective, a current correlation value could be inferred from
other traded heat rate options. If such options do not exist, then a value can be
derived using historical forward prices from the two underlying fuels. The use of
options would be preferable as the value derived would be an implied, forward-looking
correlation figure; the use of forwards returns a historical value. Typically, longdated correlation figures tend to be relatively high because the two fuels will be
impacted by similar fundamentals. The correlation for shorter-dated transactions
can be lower as physical issues such as transmission problems and weather will
impact prices.
Like all other options, there are several Greeks that a trader will need to manage. With respect to the delta, this could be hedged by trading the underlying forwards
(‘swaps’). Greeks such as vega, gamma, and theta can be hedged using other heat rate
options, but it is also possible to use individual options on gas or electricity if the trader
believes they offer greater liquidity.
CHAPTER
9
Plastics
9.1
THE CHEMISTRY OF PLASTIC
‘Plastic’ is a general term used to describe different chemical structures. At a very basic
level, plastics are formed when carbon and hydrogen atoms are combined in different
ways. Groups of atoms bonded together form molecules which, when linked together
in different ways, yield plastics that have different physical properties.
Hydrocarbons (i.e. molecules of hydrogen and carbon) are classified as either
monomers or polymers. A monomer (mono = one) is a building block molecule, which
could be chemically reacted to make molecules with longer chains. A polymer (poly
= many) is a number of individual monomers chemically joined by a bond to form
a single structure. For example, the simplest of plastics is polyethylene (sometimes
referred to as polythene), which is made up of ethylene/ethene1 monomers (Figure 9.1).
Chemically, the ethylene monomer is expressed as:
H
H
C
C
H
H
FIGURE 9.1 Chemical structure
of an ethylene (a monomer).
Where:
C = Carbon atoms
H = Hydrogen atoms
An ethylene molecule has two carbon atoms, which are joined by a double bond.
Double bonds are relatively easy to break allowing a new atom or molecule to join the
original structure.
1
Ethylene and ethene are different terms used to describe the same chemical structure. Differences in their usage may occur on a geographical basis. The slightly old fashion term of ethylene
will be used throughout the text.
332
333
Plastics
When it is chemically reacted with other ethylene molecules a polymer chain
is formed to make polyethylene (Figure 9.2). Chemically polyethylene is written as
[C2 H4 ]n where the subscript n is used to denote an integer that determines the length
of the polymer chain.
*
H
H
C
C
H
H
*
n
FIGURE 9.2 Chemical structure
of polyethylene (a polymer).
Where:
C = Carbon atoms
H = Hydrogen atoms
n = an integer that determines the length of the polymer
Although hydrogen and carbon are the two main building blocks, other elements
(such as fluorine, chlorine, iodine, and bromine) can be added to the monomer
conversion process to change the physical characteristics of the resultant polymer.
For example, at the monomer level if chlorine were to replace one of the hydrogen
atoms in ethylene, the result is chloroethene (vinyl chloride), which is C2 H3 Cl. After
polymerisation, it will form Polyvinyl Chloride or PVC, which has a higher tensile
strength but a lower melting point than polyethylene.
Because there are thousands of different hydrocarbons it is often convenient to
divide them into categories. One such category is alkenes of which ethylene is an
example. Alkenes generally have a simple chemical structure, are cheap to make, and
are relatively easy to polymerise. Alkenes are also sometimes referred to as ‘olefins’
and their polymers (such as polythene and polypropylene) ‘polyolefins’.
9.2
THE PRODUCTION OF PLASTIC
In the previous section we described the chemical composition of polyethylene and its
main monomer building block, ethylene. Here we will describe the main sources of
ethylene and how plastic is produced.
Plastics can be manufactured from a variety of hydrocarbons, which are extracted
from either crude oil or natural gas. There are a variety of hydrocarbons that can be used
in plastic production which include:
▪ Ethylene – C2 H4
▪ Propylene – C3 H6
▪ Butene – C4 H8
The most common building block is the monomer ethylene introduced in the previous section, which can be derived from either crude oil or natural gas.
334
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
9.3
MONOMER PRODUCTION
9.3.1
Crude Oil
Crude oil has very limited applications by itself and so is refined to produce a variety
of products. Chemically, crude oil is made up of many different hydrocarbon structures
that are then separated (using a process known as fractional distillation) into different
components denoted by the length of their carbon chains. Examples of refined short
chain molecules are ethene, propene, and butene, which are also monomers that can
be used to make plastic. Longer chain carbon molecules such as naphtha can also be
used as part of the plastic production process. Naphtha can then be broken up (using
a process known as steam cracking) to yield a number of different products such as
high-grade petrol and ethylene.
9.3.2
Natural Gas
Once extracted from the ground natural gas will be processed to remove any impurities
and then separated into its component elements, referred to collectively as Natural Gas
Liquids (NGLs). These include:
▪ Ethane
▪ Propane
▪ Methane
▪ Butane
▪ Pentane
Once the ethane (C2 H6 ) is collected it is steam-cracked to produce ethylene. This is
then polymerised to become polyethylene, which is sometimes referred to as polyethene
or simply polythene.
However, there is a trade-off between the two routes to produce ethylene. Producing
ethylene from natural gas is cheaper but yields a smaller amount; producing ethylene
from crude oil is more expensive but returns a higher yield. Geographical access to the
raw materials has also played a role. Where low cost access to natural gas is possible it
is usually the feedstock of choice for the production of ethylene and polyethylene.
9.4
POLYMERISATION
The next step is polymerisation, which involves reacting ethylene with a catalyst. The
choice of the catalyst at this point will impact the final polymer produced as it restructures the bonds that link the carbon and hydrogen atoms. The resultant polymer will
be in the form of:
▪ Pellets
▪ Film
▪ Resin
▪ Powder
Plastics
9.5
335
APPLICATIONS OF PLASTICS
Once a particular polymer has been made it has to be fabricated into a final usable
product for purchase by a consumer. Listed below are a number of well-known polymers
with examples of some of their day-to-day applications:
▪ Polyethylene (PE) – Low Density Polyethylene (LDPE) is used for food bags and
squeeze bottles. High Density Polyethylene (HDPE) is used for detergent bottles,
refuse bags, and shopping carrier bags. The term linear simply refers to the fact
that the carbon atoms are ordered in a straight line, which results in a polymer that
takes up less physical space.
▪ Polypropylene (PP) – when rigid it can be used for caps and other closures, when in
flexible form used for films for confectionary and tobacco.
▪ Polyethylene Terephthalate (PET) – used for soft drink containers, squeeze bottles
and oven safe food trays.
▪ Polyvinyl Chloride (PVC) – thin films, clothing, bottles, and cartons.
▪ Polystyrene (PS) – compact disc cases, cartons, and medicine bottles.
▪ Polyterafluoroethylene (PTFE) – which has non-stick applications used for such
items as kitchenware.
Once formed, a polymer is said to be thermoset or thermoplastic. The term ‘thermo’
refers to the effect of heat while the terms ‘set’ and ‘plastic’ refer to how the polymer
reacts to the heat. A polymer described as a thermoplastic is one where the application
of heat will cause the product to change shape, but will not lead to any change in its
chemical composition. This would allow the material to be remoulded back into its
original shape or may allow for recycling. Examples of this are polythene and nylon. A
polymer that is described as being a thermoset is where the application of heat alters
not only the shape but also its chemical composition. This means that once it has been
melted it cannot be remoulded and will solidify irreversibly – it is ‘set’. Examples in this
category include polystyrene and PTFE.
Thermoplastics, which constitute the greater demand of the two categories, can be
converted into a final product using a number of processes that include:
▪ Injection moulding – the plastic is fed into a heated chamber whereupon it softens
into a fluid. It is then injected into a mould and cooled so it solidifies. The mould
is then removed, and the object can be removed. This technique is used in the production of butter tubs and yoghurt containers.
▪ Blow moulding – here molten plastic in the shape of a tube is formed. Using compressed air, the tube is then blown to fill the insides of a chilled mould. This method
is used to make items such as bottles or tubes.
▪ Extrusion – in this process the plastic is heated in a chamber and then when in
molten form is pushed through an opening called a die. The shape of the opening
will allow the plastic to cool and set in a particular form.
Some polymers may undergo further processing once they are formed to give them
specific qualities that are suited to specific jobs. For example, rubber can be vulcanised
(i.e. heated with sulfur) to make it stronger and more resistant to heat.
336
9.6
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
SUMMARY OF THE PLASTICS SUPPLY CHAIN
When considering the structure of the plastic production process, broadly speaking
there are two different types of producer. An example of an integrated producer would
be a large oil major that can extract crude oil or natural gas and then send it for refining.
After refining the manufacture of monomers and polymers will follow. A nonintegrated producer would be a buyer of the monomer and would then produce a
particular type of polymer.
From that point a converter will purchase a particular polymer either directly from
a producer or from a distributor and then manufacture an end product for use by an end
consumer.
Traders will act as a counterparty to different entities in the physical supply chain
to facilitate the purchase and sale of different polymers and to offer a variety of risk
management services.
9.7
PRICE DETERMINATION
In May 2005, the London Metal Exchange (LME) launched two futures contracts in
an attempt to make plastics a tradable commodity. The LME initially listed two different types of contract, polypropylene (PP) and linear low-density polyethylene (LDPE)
since at the time these two commodities accounted for about 40% of the thermoplastics
market. In the years that followed the LME made a number of changes to the contract
to increase the degree of interest from participants along the supply chain. However,
the contract was delisted in April 2011 as the exchange was unable to establish any
significant volume or open interest.
Recall that a futures exchange has three principal roles:
1. A central reference point for the pricing of commercial contracts for the underlying
commodity.
2. Provides a series of instruments that allow participants to hedge an underlying
exposure.
3. Provides participants with a source of supply for the underlying commodity or
allows them to dispose of any excess inventory.
One feature of commodity products where no traded market exists is the absence of
a transparent method for pricing commercial contracts. Plastics are an interesting case
study as they illustrate the role of benchmark indices and their adoption by the market.
For an industry to accept a benchmark index it must be:
▪ Transparent,
▪ Representative of the most liquid market,
▪ Used by a high number of varied market participants along the supply chain,
▪ Based on actual trades executed rather than prices quoted.
Since the demise of the plastics futures, commercial transactions have been valued
using indices such as IHS Markit and PetroChem Wire (PCW).
Plastics
9.8
337
PLASTIC PRICE DRIVERS
In many respects, the market factors that influence the price of plastics share much in
common with those that influence commodities generally. Since crude oil and natural
gas play a significant part in the production of plastic, the price factors that influence
these commodities will also be important. The key price drivers for the market include:
Cost of crude oil and natural gas – Since the basic feedstock in the production process is crude oil or natural gas, the final price of plastic will be heavily influenced by
price movements in these sectors.
Environmental concerns – Very often the proposed construction of a new petrochemical site in a particular location raises issues surrounding the impact on the environment. This may delay the building of a new facility and add further to any existing
capacity constraint.
Recycling – Traditionally, plastics have not been easy to recycle, as they are often
complex and composite materials. However, investment in new technologies has
increased as a result of public concern over single-use plastics and the perceived
pollution. For example, in 2015 the UK introduced a small nominal charge of 5p for
disposable grocery-style plastic bags, which led to a decline in their usage of about
90%. According to The Independent (2019) the average English shopper now uses
just 10 bags a year compared to 140 before the charge came in. For plastics that can
be recycled, the bags will act as an extra source of supply. Typically, plastic waste is
collected from households and recycling bins. It is then taken to a recycling centre
where the plastic is sorted into different grades. Each type of plastic is then washed and
decontaminated. The plastic is then shredded into flakes which are heated and turned
into pellets. These pellets are then sold to manufacturers to make new containers.
Production capacity – If a commodity experiences a sharp increase in price, economic theory tells us that production should respond accordingly. What this does not
take into account is the fact there may often be a substantial time lag if there is a need
to expand any existing facilities. Lack of production capacity could include not only
refining of crude oil or natural gas but also polymerisation.
One concept that links some of the price factors together is the regional cost of the
feedstock relative to the available capacity. Since hydrocarbons are cheaper to produce
in the Middle East, there is expected to be a gradual trend to locate new refining and
polymerisation plants in this region. Given the difficulty of building the equivalent facilities in the West it would seem that reliance on the Middle East for energy is only likely
to increase.
Production disruption – This could be due to a variety of reasons such as natural
disasters or labour disputes. For example, hurricanes on the US Gulf Coast have on
occasion caused the shutdown of production facilities with US demand being met by
imports from other regions such as Asia.
Economic cycle – If the economy is experiencing a boom period then there will be a
general increase in production with an associated uplift in the demand for plastics.
Level of existing inventories – If prices are rising due to increased demand and production has been unable to respond immediately the excess has to be met through
existing inventories.
338
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Technological progress – As technological improvements have taken hold, the possibility of substituting plastic for other materials has increased. For example, in the
beverage industry, cans have gradually been replaced by PET containers, which have
become lighter, stronger and more tolerant to heat.
Strength of the US dollar – Similar to most commodities, plastics are quoted in
USD. A strengthening dollar will make plastics relatively more expensive in domestic
non-dollar currency terms and vice versa.
Emerging markets demand – Again similar to most metals, the rapid economic
expansion and the associated increase in demand of countries such as China and India
will have a significant bearing on the price of the commodity.
9.9
FORWARDS AND SWAPS
The process of hedging allows an entity to protect the economic value of some underlying exposure.
Section 9.6 outlined a simplified physical supply chain for the plastics industry. In
the absence of a liquid futures market it is still possible to trade OTC forwards, swaps,
and options, which can be used at each stage of the lifecycle. Table 9.1 provides a recap
of the supply chain, the nature of the price exposure and whether the participant is
likely to be a forward buyer or seller.
Reference is sometimes made to ‘price fixing hedges’ and ‘offset hedges’. A price
fixing hedge will be a transaction that is not linked to a specific underlying exposure
and is designed to alter the overall ‘macro’ price risk of a commercial enterprise. On the
other hand, an offset hedge will be based on a specific transaction and can be thought
of as a micro hedge.
Price fixing hedge
The manufacturer of any product may be faced with a situation where the required
raw materials may not be priced until their actual receipt. This means the company is
exposed to a possible change in its profit margins. As a pre-emptive action, the manufacturer could buy on a forward basis using a contract whose maturity coincides with
the pricing date for the underlying commodity. Since the forward is not being used as a
source of supply, the derivative transaction could be closed out by taking an equal and
opposite position shortly prior to maturity.
Say that a manufacturer of plastic water bottles has an ongoing need to buy linear
low-density polyethylene. He does not feel he is able to pass on any extra costs to the
end user of the water bottles and so decides he need to lock in costs in order to maintain
his margins.
Looking at the prices in early March, he notes that a June forward contract is trading
at USD 1,185/tonne and so decides to buy one forward at this price. Let us assume that
the manufacturer decides to close out the position at the end of May when the price for
June delivery is at USD 1,250/tonne. Having bought the forward at USD 1,185 and sold
it at USD 1,250, the manufacturer has made a profit of USD 65.00/tonne. If the purchase
of the physical product had also been based on the same agreed index price then as of
339
Plastics
TABLE 9.1 Price exposure and futures strategies for the plastics supply chain.
Supply chain role
Polymer price exposure
Forward transaction
Polymer producer
Converter
Distributor
End consumers
Traders
Falling prices
Rising prices
Falling & rising prices
Rising prices
Falling and rising prices
Sell forward
Buy forward
Sell/buy forward
Buy forward
Sell/buy forward
the end of May the net cost to the manufacturer would be USD 1,250 less the USD 65.00
dollars profit per tonne on the forward. That gives a net cost of USD 1,185, which is
equal to the price of the original purchased forward.
The hedger can also simultaneously execute a series of forward contracts for
delivery in sequential months, if they know there will be a regular underlying physical
exposure. This is described as a strip of contracts. An alternative to a strip would be
to enter a classic fixed-float swap. If the manufacturer is buying the raw materials
and is paying the prevailing market price, then a multi-period swap contract would
see them receiving floating and paying fixed. The floating leg of the swap will offset
their floating payment made to purchase the underlying commodity and they would
then be left with a net fixed cost courtesy of the fixed leg of the swap. Some market
participants would describe this as ‘buying’ the swap, which means paying the fixed
rate. The author is not a great fan of this terminology as it can be somewhat ambiguous
and is often not properly defined in customer confirmations.
Offset hedge
An offset hedge is where the forward is used to protect the value of a specific transaction.
Let us assume that a distributor has agreed to sell some PP in two months’ time. The
distributor has agreed a fixed sale price with the end customer. The distributor is not
obliged to hedge this exposure but runs the risk that the cost of the raw material will
rise prior to the promised delivery, eroding any profit margin. As a result, they decide
to buy a forward at the current price. As long as the forward can be traded at a price
less than the fixed sale price with the end customer, they will be able to lock in their
margins.
Proxy hedges
In the absence of a liquid market for hedging, some alternative markets could be used as
proxy hedges. Suppose that a clothing manufacturer buys a large amount of polypropylene each year and is looking to hedge their costs. In the absence of a liquid market
for the underlying, it may be possible to look at alternative commodities that might
offer a partial solution. For example, if it was found that natural gas or crude oil prices
displayed a significant correlation with polypropylene, it may be possible to structure
a hedge solution using these related commodities. However, it is important to realise
340
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
that the concept of a ‘perfect hedge’ does not exist and it is possible that any historical
correlation relationship may break down when markets are stressed.
9.10
OPTION STRATEGIES
Recall one of the earlier comments made about derivative valuation: ‘if you can hedge
it, you can price it’! That is, the cost of any derivative is driven by the cost of hedging
any underlying exposure.
At the time of the launch of plastics contracts, the LME did not implement an option
contract. Their argument was that they wanted liquidity to develop in the underlying
futures market such that an option trader would have the means to hedge their directional option exposure. Since the futures market never matured, options on this asset
class will likely remain relatively illiquid.
The main option exposures are ‘delta’ (the direction of the underlying price) and
‘vega’ (the implied volatility). One way of hedging these exposures in the absence of a
liquid futures market is by the use of proxy hedges. For example, if the trader believed
that there was relatively stable correlation between natural gas and oil with the underlying plastic exposure they are looking to hedge, they could use futures and options on
these assets as an alternative hedging mechanism. It will not be a perfect solution but
in reality, it is rare to find a situation where a hedge will be 100% perfect.
Rather than repeat material covered elsewhere, the interested reader is referred to
the gold and base metals chapters for a discussion of different over-the-counter option
strategies. The chapter on gold reviews option strategies from a producer perspective,
while the section on base metals focuses on consumer transactions.
CHAPTER
10
Bulk Commodities
T
his chapter brings together several related markets, which are sometimes collectively referred to as ‘bulk commodities’. Typically, these products (e.g. coking coal
and iron ore) have a unifying characteristic in that they form part of the steel supply
chain. In the freight market, about 60% of Capesize vessels are used to transport iron ore.
The Singapore futures exchange (2019) claims:
▪ In tonnage terms, the global steel market is about 20 times larger than all other
metal markets combined.
▪ Approximately 98% of iron ore is used in the production of steel.
▪ 15% of global coal consumption is used in steel making in the form of coking coal.
▪ China makes up about two-thirds of seaborne iron ore demand.
▪ Most iron-ore reserves are found in Australia, Brazil, and Russia.
Market participants do not always accept this definition of bulk commodities. For
example, institutions that have a focus on consumers, their iron ore desks may well sit
within the base metals function. But for an entity that serves producers, putting coal,
iron, and freight into a single unit may make sense.
This chapter considers the markets for coal, iron ore, and freight. Steel markets are
described in Section 5.4. Bulk markets can also include alumina, bauxite, and grains,
but these topics are considered elsewhere.
10.1
THE BASICS OF COAL
‘Coal’, like crude oil, is a general term that is used to describe different types of a particular commodity. The World Coal Institute (WCI) defines coal as: ‘ . . . a combustible,
sedimentary organic rock, which is composed mainly of carbon, hydrogen and oxygen. It is
formed from vegetation which has been consolidated between other rock strata and altered
by the combined effects of pressure and heat over millions of years to form coal seams’.
The International Energy Agency (2006) defines coal in a different manner, categorising it as either hard coal or brown coal. ‘Hard coal with a high thermal value . . .
is economically suited to international trade, with characteristics making some coals
341
342
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
suitable for metallurgical (coking) uses. Brown coal (lignite) has a much lower thermal
value . . . and is suitable mainly for power generation locally or to a lesser extent for
briquette manufacture’.
The type of coal recovered depends upon the length of time that it has spent in formation, as well as the temperature and pressure to which it has been subjected. Often
coal is classified as being either low or high rank, each of which will have different properties. High rank coals encompass anthracite and bituminous coals. These coals will
have a higher energy and lower moisture content. Domestic and industrial consumers
use anthracite, which makes up about 1% of the world’s reserves, primarily for heating
purposes. Bituminous coals (52% of the world’s reserves) can be broken into two main
types: thermal or ‘steam coal’ and metallurgical or ‘coking coal’. Thermal coals are used
in power generation and cement manufacture, while metallurgical coals are typically
used in the manufacture of iron and steel. Coking coal is used in coke ovens to produce
coke, which is used in blast furnaces to produce pig iron. Coking coal is harder than
thermal coal and consequently trades at a premium.
Lower rank coals include lignite (‘red coal’) and sub-bituminous coals. They have
a lower energy and higher moisture content. Applications of lignite include power generation, while sub-bituminous coals are used in power generation and cement manufacture. This breakdown is shown diagrammatically in Figure 10.1.
Lower
Carbon/energy content of coal
Higher
Higher
Moisture content of coal
Lower
Coal type
Low rank coals
(47%)
Lignite
(17%)
Power generation
Hard Coal (53%)
Sub-bituminous
(30%)
Power generation
Cement manufacture
Industrial uses
Bituminous (52%)
Thermal
(Steam Coal)
Power generation
Cement manufacture
Industrial uses
FIGURE 10.1 Different types of coal and their applications.
Source: Author, World Coal Association
Anthracite (~1%)
Metallurgical
(Coking coal)
Manufacture of iron
and steel
Domestic/
industrial
343
Bulk Commodities
10.2
THE DEMAND FOR AND SUPPLY OF COAL
Despite environmental concerns, coal is still a popular source of energy. Figure 6.1
showed the primary energy supply by fuel as of the end of 2019 and illustrates that
coal accounts for 27% of consumption, second only to crude oil. According to the World
Coal Association:
▪ 41% of global electricity is fueled by coal-fired power plants.
▪ 70% of all steel production uses coal.
▪ 200 kg of coal is needed to produce one tonne of cement.
▪ 50% of the energy used to produce aluminium comes from coal.
The following chart (Figure 10.2) shows the proportion of proved reserves for coal
by region.
BP defines the R/P ratio (reserves to production ratio) as reserves remaining at the
end of any year divided by the production in that year. The result being the length of
time that those remaining reserves would last if production were to continue at that
rate. Regional R/P ratios are shown in Figure 10.3.
Figures are for commercial solid fuels only, i.e. bituminous coal and anthracite
(hard coal) and lignite and brown (sub-bituminous) coal.
Figure 10.4 shows how the nature of coal production has evolved since 1981 indicating its ongoing popularity.
North America
24.10%
Asia Pacific
42.70%
S & Cent. America
1.30%
Europe
12.60%
Middle East and Africa
1.50%
CIS
17.80%
FIGURE 10.2 Regional coal reserves.
Source: BP Statistical Review of World Energy 2020. BP PLC.
344
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Middle East and Africa
57
Asia Pacific
77
S. and Cent. America
152
Europe
244
CIS
338
North America
367
0
50
100
150
200
250
300
350
400
FIGURE 10.3 Reserves to production ratio. Horizontal axis is years.
Source: BP Statistical Review of World Energy 2020. BP PLC.
180
S & C America
Africa
Europe
CIS
N America
160
140
120
100
80
60
Asia Pacific
40
20
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
0
FIGURE 10.4 Regional production of coal in exajoules, 1981–2019. Middle East is not included
due to relative size.
Source: BP Statistical Review of World Energy 2020. BP PLC.
Figure 10.5 shows the regional consumption of coal expressed in million tonnes of
oil equivalent.
According to the IEA (2019) the four largest exporters of thermal coal (in descending order) are:
▪ Indonesia (435 MT)
▪ Australia (203 MT)
▪ Russia (173 MT)
▪ Colombia (81 MT)
While the four largest exporters of coking coal (in descending order) are:
▪ Australia (179 MT)
▪ USA (56 MT)
345
Bulk Commodities
180
S & C America
Africa
Europe
CIS
160
140
120
N America
100
80
60
40
Asia Pacific
20
19
65
19
67
19
69
19
71
19
73
19
75
19
77
19
79
19
81
19
83
19
85
19
87
19
89
19
91
19
93
19
95
19
97
19
99
20
01
20
03
20
05
20
07
20
09
20
11
20
13
20
15
20
17
20
19
0
FIGURE 10.5 Regional consumption of coal in exajoules. Middle East is not included due to
relative size.
Source: BP Statistical Review of World Energy 2020. BP PLC.
▪ Canada (29 MT)
▪ Russia (26 MT)
The main destinations for both types of coal are:
▪ China (295 MT)
▪ India (240 MT)
▪ Japan (185 MT)
▪ South Korea (142 MT)
▪ Chinese Taipei (66 MT)
Figure 10.6 illustrates the movement of the price of steam coal over the period 1983
to 2005 for five major markets (average export unit values in USD per tonne).
250.00
200.00
150.00
100.00
50.00
0.00
1987
1989
1991
1993
NWE
1995
1997
US
1999
2001
2003
Japan – coking
2005
2007
2009
Japan – steam
FIGURE 10.6 Coal prices 1987–2019.
Source: BP Statistical Review of World Energy 2020. BP PLC.
2011
2013
2015
Asian marker
2017
2019
346
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The growth in the production and consumption of coal has been fueled by the following factors:
▪ Increased demand for energy globally.
▪ In some regions coal is cheaper per kilowatt-hour than competing energy sources
such as natural gas and crude oil.
▪ The fuel is relatively safe to transport and costs less to move than oil or natural gas.
▪ There are substantial reserves available. According to the BP Statistical Review of
World Energy 2020, coal had a reserve to production ratio of 132 years, compared
to 49.9 and 49.8 years for crude oil and natural gas, respectively.
10.3
COAL – THE PHYSICAL SUPPLY CHAIN
10.3.1
Production
The two methods of extracting coal are by underground or surface (open cast) mining.
Although more coal can be recovered from a given deposit by surface mining, most of
the coal is extracted from underground mines. Like most commodities that are extracted
from the ground it will not be in a form that will be immediately usable. Depending on
the physical form of the coal upon its extraction, it may require crushing to make the
product physically manageable. The coal is also cleaned and will have any impurities
removed.
Inevitably the coal will have to be transported to its final location, which could
involve a variety of different transport modes depending on the distance to be travelled
and the final delivery location. These include:
▪ Lorries
▪ Trains
▪ Barges
▪ Ships
10.3.2
Main participants
The main participants in the coal supply chain and their associated risk management
concerns include:
Mining companies – Large companies will use risk management solutions to add
value to their operations and will tend to trade in significant volumes. Other mines may
be looking for financing solutions often linked to coal prices and FX movements.
Utilities – These tend to be power-generating companies that operate coal-fired
power stations. They may have a preference to use financially settled derivative transactions so they can separate out the physical delivery from the associated price risk, which
can exhibit substantial volatility. With the increasing concern over emissions trading
they must also consider the cost of emitting carbon.
Non-physical participants (e.g. banks and trading houses) – Their trading activity
with other financial desks creates liquidity for the market, thereby facilitating the development of risk management solutions. Equally, banks that have a trading capability will
Bulk Commodities
347
be able to offer a hedging solution to other lending banks that wish to offer risk management solutions to their clients without the need for a trading desk. Banks with a
trading desk can also offer a mechanism for other banks to offset the market risk inherent within the issuance of commodity-linked investment structures.
Industrial companies – Cement producers are major coal consumers, and the cost
of the raw material can contribute a significant amount to their overall costs. Steel producers use coking coal but financial markets only trade thermal coal. This is the best
available hedge but is not perfect in that the hedger has an exposure to basis risk (i.e.
the risk that the price of the two commodities does not move in the same direction).
10.3.3
Factors affecting the price of coal
Steam coal is not a homogeneous product and as such several benchmarks have evolved
based on standardised specifications surrounding the ash, sulfur, and energy content.
The main pricing locations from which indices have evolved:
▪ Central Appalachia (USA)
▪ Powder River Basin, Wyoming (USA)
▪ Europe ARA (Amsterdam, Rotterdam, Antwerp)
▪ Richards Bay (South Africa)
▪ Newcastle (Australia)
Not all these locations, however, are traded internationally.
Electricity industry – Since electricity represents one of the largest uses of coal, an
increase in the electrification of a country (e.g. India) will have an impact on price.
Although in this instance it may be tempting to suggest that the same would be true of
China, this is one instance where China has abundant reserves of coal, which it can use
to meet this demand. Planned capital expenditures, for the building of new coal-fired
power generating facilities, will influence perceptions of the future balance between
demand and supply.
Electricity production margins – The main fuels used to produce electricity are natural gas and coal. If the price of natural gas rises relative to coal, then it may encourage
producers to switch. With the advent of allowances for installations that emit carbon
dioxide, the cost of polluting now must be taken into consideration. This cost is now
also expressed in terms of the ‘dark spread’ and ‘clean dark spread’.
Technology – Because of the increased environmental concerns, much money has
been spent on improving the performance of coal at the point of consumption while
aiming to reduce the undesirable environmental side effects. The efforts to clean coal
either involve the way in which it is prepared prior to its consumption or by fitting
equipment to power stations to reduce undesirable emissions such as sulfur. Another
focus of the technology has been to increase the thermal efficiency of coal by fitting
more efficient boilers that reduce the carbon dioxide emissions.
Environmental issues – Coal is perceived as a significant polluter of the environment
and any associated regulation that may hinder its production or consumption could
have an adverse effect on its market price. Methane, which is formed alongside coal,
is released when the coal is mined, and is classified as a greenhouse gas. Additionally,
348
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
when coal is burnt it releases pollutants such as carbon dioxide, ash, and oxides of sulfur
and nitrogen (SOx and NOx ).
Freight costs – Although not all coal is traded internationally, the price of coal will
be influenced by the cost of transport to an end location.
Economic expansion – As an economy expands there may be an increase in the
demand for coal. Take for example, the economic expansion of China, which has led
to an increase in the demand for steel. This has led to an increase in demand for coking
coal to fuel the blast furnaces used in the production process.
Coal quality - The heat and sulfur content are considered the primary determinants
of steam coal. For example, coal mined from the Powder River Basin in Wyoming is
more popular with utilities as it has a lower level of sulfur. Combined with the fact that
it is mined from open cast pits, it is cheaper and cleaner than coal extracted from the
Appalachian Mountains.
One of the main value drivers of coal is its calorific value. Like the concept of a crude
oil assay, a sample of coal needs to be checked in laboratory conditions to ensure that it
is in line with the specifications defined in a contractual supply agreement. Although
the calorific value of coal is affected by several factors, the amount of the moisture in
the coal will impact the measure. Some coal contract specifications include the terms
‘net as received’ (NAR) and ‘gross as received’ (GAR). When coal is quoted on a GAR
basis the associated calorific value (sometimes known as the upper heating value) is
the gross calorific value obtained under laboratory conditions after all the moisture has
been removed. Quoting on a NAR basis refers to the net calorific value obtained in boiler
plants and is known as the lower heating value. The difference between the two would
be the heat of the water vapour produced.
Mining related issues – Since coal is extracted from the earth, there are several factors that could influence the price of coal. These were outlined in the chapter on base
metals and include:
▪ Capital expenditure
▪ Production disruption
▪ Production costs
Whether the coal is mined from open cast or underground sites will influence the
cost and drive the economics of production. Reserves that are deep underground are
less economical to mine.
In some countries such as Russia the coal may need to be transported over significant distances to reach the final point of consumption. Related to this is the availability
of logistics at the points of loading and discharging. Equally, since coal is often shipped
by rail, derailments and subsidence on the line may have an impact on supply.
Competing fuel costs – Liquefaction is the process whereby coal is transformed into
a liquid. The resultant liquid fuel can be refined to produce fuels or other oil-based
products such as plastic. This conversion will become more attractive as prices for crude
oil rises.
In a similar vein gasification converts coal into synthetic gas, which can then be
used by power stations to generate electricity. Coal, water, and oxygen are mixed in an
asifier, which acts like a high-pressure cooker. The combination of heat and pressure
Bulk Commodities
349
chemically breaks down the coal to produce synthetic gas, while impurities such as
sulfur are removed. This ‘syngas’ can be burnt to drive a gas turbine to produce electricity. As an undesirable side effect, the gasification process produces carbon dioxide, but
technology is evolving to capture this and bury it underground.
Although coal is still used in the production of electricity, the increase in the availability of natural gas (and its resultant price fall) could lead to some electricity generators switching away from coal. However, since natural gas is more of a regionally traded
product, an electricity generator in a country with low natural gas reserves would be
reliant on supplies of LNG.
Influence of non-physical participants – The entrance of banks, hedge funds, and
institutional investors could be significant. An increase in investment interest could
result in an upward influence on the price, while the trading activities of the banks
and hedge funds could provide enhanced liquidity with respect to the provision of risk
management structures for the physical supply chain.
10.4
COAL DERIVATIVES
Estimates of the size of the coal market vary but one crude method of assigning a monetary value is to take the volume produced each year and multiply it by the prevailing
market price. Using a 2019 production figure of 8.1 billion tonnes1 and a market price
of about USD 60.00/MT, this would imply a value of USD 486 billion.
The derivatives market for coal is a fraction of this amount with notional values
totaling about USD 100 billion. In most derivative markets, this figure would be a multiple of the physical market, suggesting that the coal market is relatively illiquid.
Most transactions are financial swaps, but vanilla options are also traded. Financial
Coal swaps have been trading in the OTC markets for about fifteen years. There are
several different indices traded in the markets reflecting the key locations for seaborne
thermal coal. The locations are either sources of coal (South Africa and Australia) or a
consuming location such as Europe.
Examples of popular indices include:
▪ API2 – This is a CIF (cost, insurance, freight) price for delivery in Rotterdam, which
is an average of certain prices and is published by Argus and IHS McCloskey. This
is a benchmark for the Atlantic coal market.
▪ API4 – This is a FOB (free on board) price for delivery from Richards Bay, South
Africa. This index derives its value from prices quoted by Argus and IHS McCloskey.
This is a benchmark for the Pacific coal market.
▪ GlobalCoal NEWC index – this price relates to coal originating in Newcastle, Australia.
Within the context of the coal market, API does not refer to the density of the
material but simply stands for average price index. This reflects that the index is compiled from several different observations and the final price is a simple average of these
1
BP Statistical Review of World Energy 2020.
350
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
prices. The number (i.e. 2 and 4) that follows the acronym has no significance apart
from allowing practitioners to distinguish between the two different pricing locations.
Argus/McCloskey quote prices for API 2 (Northwest Europe), 3 and 4 (South Africa),
5 and 6 (Australia), 8 (South China), 10 (Colombia), and 12 (East India). Where two
indices exist for a single geographical location the main differentiator will be one of
quality; API 3 is for lower grade coal exported from Richards Bay than API 4.
Driven by the liberalisation in the electricity markets across Europe, the utilities,
who are the biggest users of financial swaps on coal, have been faced with less predictable load factors and are, therefore, moving away from using the one-year fixed
price supply contracts, which were the norm. This has led to an increase in hedging
requirements using price indices.
Prices for coal are considered to be relatively volatile, but still are at the lower end
of competing fuel sources. The following ranges of implied volatilities are indicative:
API2 Rotterdam
API4 Richards Bay
GlobalCoal NEWC
Crude Oil
Natural Gas (US)
Natural Gas (EUR)
10.4.1
25–35%
20–25%
18–25%
25–50%
40–120%
40–90%
Exchange traded futures
Exchange traded coal futures exist, but at the time of writing are somewhat illiquid. One
example is the API2 Rotterdam coal futures traded on ICE. The contract specification
is detailed below (Table 10.1):
TABLE 10.1 Coal Futures Specification.
Trading Unit
1,000 metric tons of coal
Price quotation
Trading month
US dollars and cents per ton
The exchange trades individual month, quarterly,
seasons, and annual contracts
USD 0.05 (5 cent) per ton
Trading terminates on the day of expiry of the delivery
month, quarter, season, or calendar. So, a monthly
contract of, say, March will terminate on the last
business day of that month.
Cash settled at an amount equal to the monthly average
API2 index as published by Argus McCloskey’s coal
price index report
Minimum price fluctuation
Last trading day
Delivery
Source: Data from API2 Rotterdam Coal Futures, ICE FUTURES EUROPE, Intercontinental Exchange, Inc.
Bulk Commodities
10.4.2
Over the counter solutions
10.4.2.1
Coal swaps
351
Arguably, the most popular OTC solutions in the coal markets are swaps and swaptions
on coal. Take for example the risk profile of a large Australian coal producer. The company sells most of its production through traditional long-term fixed price contracts,
with the balance sold against an index (API 2 for delivery in Rotterdam) on a floating
basis. They have chosen this index pricing basis for their exports as they have access
to very favourable shipping rates. However, using a floating index gives the company
exposure to fall in the price of coal. As part of its risk management strategy it is willing to swap this index exposure into a fixed price if it believes that the ‘price’ (i.e. the
fixed element of the swap) is at an attractive level. It decides to enter a swap with the
following terms:
Trade date:
Maturity:
Total notional quantity:
Notional quantity per month:
Settlement dates:
Payment date:
Fixed price payer:
Fixed price:
Floating rate payer:
Reference floating price:
Tenor:
October
Three calendar months, effective from the
following June.
15,000 MT (metric tonnes)
5,000 MT
Monthly in arrears
Five business days after settlement date
Bank
USD 60.00/MT
Client
COAL – API 2 – ARGUS / MCCLOSKEY’S
Monthly price
With the company receiving the floating index on their underlying physical exposure but paying the same index under the terms of the swap, the net effect is that the
two floating exposures cancel leaving the producer receiving a fixed cash flow of USD
60.00 per metric tonne. This presumes, however, that the date at which the floating
price is fixed for the swap coincides with that of the physical exposure. A mismatch in
the timings of the floating payment would be referred to as basis risk.
In this scenario the producer sells most of their production under long-term fixed
price agreements. If the demand for coal were to increase, the producer would still use
swaps to benefit from any rise in price. They could enter a coal swap referencing the
fixed price component of their production. Under the terms of the swap they would
receive a floating index price and pay fixed. The net effect of this would have been to
leave the producer as a net receiver of the floating index, allowing them to benefit from
the rise in the price of coal.
In the previous example the cash flow exchanges took place monthly. This would be
useful as a hedge for a producer who is selling a regular amount of coal per month over
the period in question. However, the swap could also be structured as one three-month
contract. Here the contract is most likely acting as a cash-settled forward transaction,
where the price is fixed at the start of the period but settled in arrears. For example:
352
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Trade date:
Maturity:
Total notional quantity:
Notional quantity per period:
Settlement dates:
Payment date:
Fixed price payer:
Fixed price:
Floating rate payer:
Reference floating price:
Publication source:
Tenor:
October
Three months effective the following June
15,000 MT (metric tonnes)
15,000 MT
One three-month period paid in arrears
Five business days after settlement date
Bank
USD 60.00
Client
COAL – API 2 – ARGUS / MCCLOSKEY’S
Argus /McCloskey’s coal price index report
Three-month price at start of period
A common feature of OTC coal swaps is that the agreements will include a currency
conversion provision, which allows for the payment of cash flows to be denominated in
a currency other than USD. For example, a contract that makes the cash flow payments
in either GBP or EUR may be based on the average exchange rate for the settlement
period (e.g. monthly as in the previous example).
10.4.2.2
Coal swaptions
Another popular trade in the market is the coal swaption (e.g. an option on a coal swap).
A possible term sheet for such a trade might look as follows:
Option trades
Trade date:
Expiry:
Option style:
Option seller:
Option buyer:
Premium per MT:
Total premium:
Underlying transaction
Effective date:
Maturity:
Total notional quantity:
Notional quantity per period:
Settlement dates:
Payment date:
Fixed price payer:
Fixed price:
Floating rate payer:
Reference floating price:
Tenor:
September
December (three months)
European
Bank
Client
USD 1.85
USD 27,750
December, 12 months after option expiry
Three months
15,000 MT (metric tonnes)
5,000 MT
Each month in arrears
Five business days after settlement date
Bank
USD 60.00
Client
COAL – API 2 – ARGUS / MCCLOSKEY’S
Monthly price
Bulk Commodities
353
Motivation for using swaptions
The rationale for the trade will be different depending on the end client. If the client
is a producer or consumer of coal, then it is most likely that the motivation will be to
hedge, on a contingent basis, some form of underlying exposure. For example, the terms
of the underlying swap in the previous term sheet are identical to those presented in the
first swap example. We had argued that in that scenario a producer selling on a variable
rate basis could receive a fixed price to transform their exposure to a falling coal price.
A possible motivation to buy this swaption is that they believe the price of coal will rise,
but cannot afford for their view of the market to be wrong.
Another popular strategy occasionally used by producers is to sell out-of-the-money
(OTM) call options. Suppose that the current price for coal (e.g. the prompt futures
price) was USD 50.00. If the producer were to sell a three-month OTM call struck at USD
55.00 then using an implied volatility of 40% this would generate a premium of about
USD 2.00/MT. If the price of coal were to fall, then the option would not be exercised and
the producer would sell their planned output at the prevailing lower price but would
retain the premium, which would help boost cash flow. A similar situation would occur
if the coal price were stable or increased by a small amount. If the price of coal were to
increase to, say, USD 60.00 the option would now be in-the-money (ITM). The option
buyer would exercise and if physically delivered would receive the agreed amount of
coal at the strike price of USD 55.00. From the producer’s perspective their net income
would be USD 57.00 – the strike price received + the premium. Indeed, once the option
is exercised this sum will be their income irrespective of the final price of the underlying.
In this sense the sale of the call option would prevent them from benefiting from a sharp
rise in the underlying price.
In some financial markets this strategy is often referred to as a ‘covered call’ and
represents a situation where the option seller is long the underlying asset and sells
optionality on this position. Traditionally, the motivation is that the strike is set at a
level that the seller believes will not trade and so they will be able to retain the premium. Since the strategy restricts the ability of the seller to benefit from a sharp increase
in price, it would not be suitable for those participants who have a particularly bullish
outlook on the underlying.
10.5
IRON ORE
10.5.1
Background
According to the US Geological Service:
‘The element iron (Fe) is one of the most abundant on earth, but it does not
occur in nature in useful metallic form. Iron ore is the term applied to a natural
iron-bearing mineral in which the content of iron is sufficient to be commercially usable. Metallic iron, from which steel is derived, must be extracted from
iron ore.’
354
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Reference is often made to ‘fines’, which are essentially heavy grains (like sand).
The fines are then transformed into pig iron, which can then be used to make steel
using coking (metallurgical) coal in a furnace. Iron ‘lump’ is an intermediate product
that sits between iron ore and pig iron.
How iron is used in the production of steel is outlined in Section 5.4.
Steel is a combination of iron along with a small amount of carbon. Thousands of
products having various chemical composition, forms, and sizes are made of iron and
steel by casting, forging, and rolling processes. Iron and steel comprise about 95% of all
the tonnage of metal produced annually in the United States and the world. On average,
iron and steel are the least expensive of the world’s metals. In some circumstances,
such as steel framing for large buildings, no other materials are deemed suitable due
to strength requirements.
According to the USGS (2020) the five largest producers of pig iron are:
▪ China 820 million metric tonnes (MMT)
▪ India 75 MMT
▪ Japan 75 MMT
▪ Russia 50 MMT
▪ Korea 48 MMT
10.5.2
Evolution of iron prices
Iron ore prices are often quoted in terms of its ferrous content expressed as a percentage.
One of the most popular grades is iron ore with 62% ferrous content. According to ICE
(2009), most commercial grades lie between 58% and 65% with key markers at 58%, 62%,
63.5% and 65%.
The evolution of iron prices is an interesting case study on how a commodity can
transition from non-traded to traded status. This was covered in Section 1.3.
10.5.3
Iron ore derivatives
At the time of writing the Singapore Exchange listed a suite of iron ore, coking coal, and
freight contracts. In relation to iron ore they were offering:
▪ Futures, swaps, and options. The options could either be options on futures or
options on swaps.
▪ The underlying commodity was primarily iron fines although a ‘lump’ contract was
also traded.
▪ The grade quality varied from 58% to 65%.
▪ The iron ore contracts are quoted on a CFR basis. CFR is ‘cost and freight’ and is
used in relation to products that are transported by sea. The seller arranges and
pays for transport to a named port. Seller will also deliver and load the goods on the
ship. However, once the goods have been loaded, the risk will transfer to the buyer.
The seller does not arrange insurance for the goods.
Bulk Commodities
355
▪ The contracts settle against an independent published index (e.g. MB iron ore index,
58% FE fines, CFR Qingdao).
▪ The contracts are cash-settled if held to maturity.
10.6
FREIGHT MARKETS – THE FUNDAMENTALS
10.6.1
Vessel types
Freight markets are divided into two main categories: wet and dry. Wet freight will
include the movement of crude oil and refined products as well as container ships, passenger liners and tugboats. Dry freight covers the shipping of raw materials such as coal,
iron ore, bauxite, and grains.
It is often convenient to categorise ships in terms of their dead weight tonnage
(dwt), which is a measure (usually expressed in metric tons) of a ship’s carrying capacity
including fuel, fresh water, crew, and provisions.
Within the dry freight market there are several different types of vessel that are
used, which include:
Handysize
These are bulk carriers of approximately 20,000–35,000 dwt and are usually
fitted with cranes.
Handymax
These are bulk carriers of approximately 36,000–49,000 dwt and may be fitted
with cranes. These vessels will carry grains and bulk commodities such as steel
products and fertilizers.
Supramax/Ultramax
About 50,000–60,000 dwt.
Panamax/Kamsarmax
These are ships in the 65,000–83,000 dwt range, but are narrower in the beam
so as to be able to navigate the Panama Canal. Their cargo is mainly coal, grain, and
‘forest products’ (e.g. packaging, recovered paper, and printing papers).
Post Panamax/Mini Cape
About 87,000–120,000 dwt.
Capesize
80,000–175,000 dwt. Cannot navigate the Panama or Suez Canal. Will transport
goods either via Cape Horn or the Cape of Good Hope. Used mostly for iron ore
delivery.
Very large Ore Carrier (Valemax, Chinamax)
About 220,000–400,000 dwt.
Within the wet freight market different cargoes will require different handling and
transport and so special types of tankers have been built to accommodate these needs.
Like the dry freight market, tankers are classified according to their capacity.
356
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Product tanker:
Panamax:
Aframax:
Suezmax:
VLCC:
ULCC:
10,000–60,000 dwt
60,000–80,000 dwt
80,000–120,000 dwt
120,000–160,000 dwt
240,000–320,000 dwt
320,000–550,000 dwt
VLCC is a very large crude carrier, while ULCC stands for ultra large crude carrier.
10.6.2
Freight charges
Within the dry freight market, a distinction is made between a voyage charter (VC) and
a time charter (TC; sometimes referred to as a trip charter). A VC is analogous to taking
a taxi from one point to another; you know how much the trip is going to cost you. A
VC is priced in USD per tonne. On the other hand, a TC is a little like renting a car; you
can go wherever you want, but you must return the car back to where you rented it or
face a drop-off surcharge. The hirer is responsible for petrol and tolls as well as having
to cope with traffic jams and other delays. A TC is priced in USD per day.
It is not always straightforward to compare TC and VC prices. The first thing is to
make sure that the two routes being compared are the same. To convert a TC into a VC
the first step is to calculate the number of days it will take to make the round-trip voyage
(allowing for loading and unloading). The cost of fuel will then need to be added. This
total cost can then be divided by the total tonnes carried to derive a figure that can be
used for comparison purposes.
10.6.3
Freight market participants
There are several different freight market participants:
▪ Ship builders
▪ Ship owners
▪ Ship brokers
▪ Charterers
▪ Operators
Brokers will act as intermediaries between ship owners and those entities with a
cargo that they need to move. Charterers might include mining companies, utilities,
steel mills, cement producers, or grain traders. A ship operator will assume responsibility for the operation and administration of a vessel (e.g. crewing, technical operation,
and maintenance). Some operators could charter a ship from an owner for several years
and then charter the vessel to another operator for a higher rate but a shorter period.
Although currently owned by the Singapore Exchange, the Baltic Exchange is a
membership organisation that provides an independent source of maritime market
information, which can be used to settle physical and derivative shipping contracts.
Bulk Commodities
357
The members of the exchange comprise of market participants who are active in the
dry and wet freight markets. It is the oldest shipping market in the world and can trace
its origins back to 1744. To facilitate maritime trade, the Baltic Exchange compiles and
publishes a range of benchmark indices. According to their website the indices ‘are
an assessment of the price of moving the major raw materials by sea. The indices are
based on the assessments of the cost of transporting various bulk cargoes both wet (e.g.
crude oil and oil products) and dry (e.g. coal and iron ore) made by leading shipbroking
houses around the world on a per tonne and daily hire basis.’
10.6.4
An overview of dry freight indices
The Baltic Exchange publishes several benchmarks that cover a wide variety of different
routes and vessels. These indices are designed ‘to provide an assessment of the prevailing market rate for a specified shipping route in the dry or wet bulk market’ (Baltic
Exchange, 2019). Several international shipbroking firms who have no direct financial
interest in chartering vessels make the assessments.
The Exchange acknowledges that the compilation of the index is challenging, citing
several factors:
▪ The shipping market is complex, varied, and often opaque.
▪ Shipping contracts are private bilateral contracts. This means the terms are not
standardised and will only be known to the two participants.
▪ Rates and prices agreed might be contingent on other conditions being fulfilled.
▪ Shipping markets are generally considered to be illiquid as transactions may be
relatively infrequent. As a result, prices can be very volatile. For example, the Baltic
Dry Index reached a value of 11,793 on 20 May 2008 but had fallen to just 663 by 5
December in the same year.
▪ There is no obligation on market participants to report transactions.
▪ Ships exist in a very large numbers of different types and sizes and may not match
those used in the compilation of a particular benchmark.
▪ Several other factors could influence the market value of a particular ship, which
might include maintenance of the vessel and the credit worthiness of the owner.
According to the Exchange, when making their assessments for the day the
‘panelists will take cognizance of the totality of market information known to them at
the time of reporting, making an appropriate adjustment to accord with the Baltic’s
route definitions.’ The Exchange (2019) points out ‘the Baltic has always explained that
reporting panels exist because ultimately, there is no independently verifiable ‘right’ or
‘wrong’ rate for index routes. Therefore, market levels at any particular time remain a
matter of judgement’.
One of the most popular indexes is the Baltic Dry Index (BDI), which as the name
suggests is not a price but an index value. This is a weighted average of three other
indices published by the exchange:
▪ Baltic Exchange Capesize 2014 Index (40% weight)
▪ Baltic Exchange Panamax Index (30% weight)
▪ Baltic Exchange Supramax Index (30% weight)
358
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Baltic Exchange
Dry Index
(BDI)
Baltic Exchange
Capesize 2014
Index
(BCI)
Baltic Exchange
Panamax Index
(BPI)
Baltic Exchange
Supramax Index
(BSI)
30% weight
30% weight
40% weight
FIGURE 10.7 Hierarchy of bulk indices.
The structure of the indices is shown in Figure 10.7.
Within each of these indexes a notional vessel is specified by the exchange (e.g. a
certain dead weight tonnage, age); within the Baltic Capesize 2014 index, the notional
vessel size is 180,000 dwt and has a maximum age of 10 years. Each index will also specify several pre-defined routes, often represented by a series of letters and numbers. Some
routes may be time charters, others voyage charters. For example, the Baltic Capesize
2014 index covers 12 routes, which include:
▪ C2 – the journey from Tubarao in Brazil to Rotterdam in the Netherlands, which is
used for the transportation of iron ore.
▪ C4 – the journey from Richards Bay in South Africa to Rotterdam in the Netherlands, which is used for the transportation of coal.
The panelists submit to the Baltic Exchange their freight rate assessments, and a
simple average monetary value is determined for each route. This is then multiplied by
a constant number (the multiplier) to convert the dollar amount into an index value.
These values are then summed, and this returns the overall value of the index.
To convey a sense of how the individual indices could be calculated, consider the
following example (Table 10.2), which has been simplified for ease of illustration.
TABLE 10.2 Index compilation – Day 1.
Route number
Assessed freight rate
(USD/day)
Multiplier
1
2
3
4
10,000
11,000
12,000
13,000
0.02500
0.02273
0.02083
0.01923
Contribution
to the index
250
250
250
250
Index total = 1,000
359
Bulk Commodities
To make the example work a few assumptions have been made:
▪ This is the first time that the index has been quoted.
▪ The assessed freight rates are time charter values, expressed as USD/day.
▪ The index compilers have set an initial index value of 1,000.
▪ The final index value is the sum of the individual route contributions.
▪ Each of the individual routes is deemed to contribute 25% to the overall value of the
index.
▪ The calculations have been rounded.
In this example, for each of the routes to contribute 25% to the initial value of the
index, their individual contributions must be 250. Given that the monetary assessments
are different, the multiplier for each route is set at such a level that it returns the appropriate index contribution. So, for route 4, with an assessed value of USD 13,000 and an
index contribution of 250, the multiplier would have to be 0.01923 (rounded). These
multipliers will then remain unchanged until there is a change in the composition of
the index, which could arise if the weightings are changed, or if routes are added or
removed. It is also possible to quote the index in USD terms and so on an unweighted
arithmetic basis the assessed value would be USD 11,500/day.
Consider now how the index would evolve if the following day, the assessed values
remain the same apart from route 4 which increases to USD 14,000. The new calculation
of the index is shown in Table 10.3.
Note that the multiplier for all the routes is unchanged as the assumption is that
there has been change to the index specification. The index contribution for route 4 has
increased to 269.23, which pushes the index value upwards to 1,019.23.
In cash terms the average value for this fictional index is now USD 11,750.
This calculation shows how the multiplier can also reflect the effect of a USD 1.00
change in the freight rate on the overall index. So, since the assessed freight rate has
increased by USD 1,000 then the index has increased in value by 19.23 points (USD
1,000 x 0.01923).
10.6.5
Worldscale
Rates for wet freight cargoes are often calculated using the Worldscale System. Worldscale is a large table of freight rates for different routes. The table covers over 300,000
TABLE 10.3 Index compilation – Day 2.
Route
number
1
2
3
4
Assessed freight
rate (USD/day)
Multiplier
10,000
11,000
12,000
14,000
0.02500
0.02273
0.02083
0.01923
Contribution
to the index
250
250
250
269.23
Index total = 1,019.23
360
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
voyages and offers different permutations about the number of loads a ship is carrying
and different discharge ports. The table is updated annually by the Worldscale organisation, a non-profit market association based in London and New York.
The table of rates is expressed in USD per metric tonne, with each rate being
referred to variously as WS100, 100% of Worldscale, or Worldscale ‘flat’ and is a base
value from which an actual rate is negotiated. Actual market rates are agreed to relative
to this figure and reflect prevailing demand and supply; a technique referred to as ‘points
of scale’. Suppose that two parties are negotiating a rate for the trip from Singapore to
Chiba in Japan. They agree to a contract rate of 287 Worldscale points. If WS100 were
USD 9.13/MT, this would equate to an amount of USD 26.2031 (USD 9.13 x 2.87).
At the heart of the Worldscale calculation is the principle that the scale will provide
the same net revenues per day irrespective of the voyage performed by a ‘notional’ vessel
at WS100. So, in theory, the rates are calculated such that after allowing for port costs,
fuel costs, and canal expenses the net daily revenue will be the same for all voyages.
Like dry bulk indices a notional vessel had to be established as the benchmark
allowing participants to negotiate a specific rate for their voyage. All the rate calculations are quoted in USD per tonne for a full cargo for the standard vessel based upon
a round voyage from loading port to discharging port. To convey a sense of the ‘net
revenue’ concept, the notional vessel specification includes some of the following characteristics:
Total capacity:
Average service speed:
Consumption of fuel:
What type of fuel used:
Price of fuel:
Port time:
Fixed hire element:
75,000 tonnes
14.5 knots
55 tonnes per day when steaming
Very Low Sulfur Fuel Oil (VLSFO – max 0.5% sulfur
content)
VLSFO USD 550.42 per tonne
Four days for a voyage from one loading port to one
discharging port
USD 12,000/day
If the contracting parties are using a larger vessel, they must agree between themselves a method of reflecting any additional costs.
10.6.6
Freight price drivers
Arguably the main price drivers for the freight market can be categorised under two
main headings: fleet development and general economic factors.
10.6.6.1
Fleet development
Availability of ships
The Baltic Exchange estimates that there are approximately 9,000 ships in the dry bulk
fleet. To put this into context there are about 13,500 yellow cabs in New York.
Vessels that are laid up for maintenance would also impact availability.
Bulk Commodities
361
Construction of new vessels
It is important to distinguish between the number of new vessels being ordered and
those that are close to being delivered. This is because on average it may take about 18
months to build a new vessel. If there is an increase in supply of vessels, then shipping
costs could fall.
Piracy
In the late 1990s and the early 2000s the focus on piracy was on the South China Sea
and the Straits of Malacca. Since 2005 the focus started to shift to piracy off the coast
of Somalia, in the Gulf of Aden and the wider Indian Ocean. For example, in 2018, the
International Maritime Bureau reported 223 incidents. The implication for shipping is
the possible rerouting of a vessel. For example, a voyage from the Gulf region to Rotterdam might take about 75 days. If the vessel were forced to travel via the Cape of Good
Hope, then the voyage time would be about 95 days. So not only does the cost of a single shipment increase, but freight rates may also increase as ships are now at sea for
longer periods. The owners of the cargo will also be faced with having to finance their
inventories for longer periods, which would increase their costs.
Choke points
If geopolitical tensions increase, there is always the risk of conflict between countries.
This could mean that locations such as the Straits of Hormuz may be deemed unsafe
areas for freight.
Environmental considerations
The general increase in concern over particulate emissions such as sulfur and nitrogen
oxides from ship exhaust systems have forced vessel owners to fit ‘scrubbing systems’
to remove the harmful matter.
10.6.6.2
Economic factors
The demand and supply for bulk commodities
The factors that influence the price of freight will often be the same as those that affect
the individual bulk commodities.
Availability of credit
A reduction in bank lending may reduce the number of ships being built, which would
represent a reduction in supply. During the financial crisis of 2007–2008 there was a
substantial fall in shipping rates and some buyers decided to cancel orders, forfeiting substantial deposits as they considered it a more financially viable option.
362
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Seasonality
This could take different forms perhaps in terms of weather conditions along a certain
route or demand for goods to coincide with a major holiday such as Christmas.
Cost of fuel
The Baltic Exchange points out that fuel costs make up between one-quarter and
one-third of the cost of running a vessel.
Infrastructure
If there is congestion at a particular port that leads to delays in loading and unloading
this is in effect a form of supply reduction as vessels will not be available.
Other key points include:
▪ Like other commodities that display a term structure, different maturities may be
influenced by different factors. For example, the number of ships to be delivered, or
the change in demand for a particular bulk commodity may influence short-term
prices. Longer-term prices may be influenced by environmental issues, which may
impact the long-term demand for coal.
▪ Some practitioners suggest that freight prices are a barometer of the current health
of the world economy, but some might argue that this is somewhat simplistic.
10.6.7
Freight Derivatives
10.6.7.1
Forward Freight Agreements (FFAs)
FFAs are essentially a form of swap and have proven to be a very popular instrument
in the shipping market. Consider a mining company that is looking to fix the price
of freight for a future date with a ship operator. This deal is sometimes referred to as
the ‘physical’ transaction. FFAs sit in the ‘financial’ or ‘paper’ side of the market and
can be used to hedge an existing freight agreement. They can also be used to express
a view on how markets are expected to evolve and do not require the participants to
have an underlying physical exposure. Those participants looking to hedge against (or
profit from) a rising rate should ‘buy’ the contract (i.e pay fixed); those looking to hedge
against falling rates should ‘sell’ the contract (i.e receive fixed). All the contracts are
cash-settled; a ULCC is not going to turn up at your office!
From a price formation point of view, in the physical market, the cost of hiring a
vessel for three months in six months’ time should be the same as the price two participants could agree in the FFA market for the same period. However, there are occasions
where differences between these two prices may emerge, which could be caused by, for
example, speculative activity in the FFA market. As such these discrepancies may offer
an attractive hedging opportunity as the transaction is trading away from ‘fair value’.
Consider the following hypothetical term sheets. Although sometimes referred to
as ‘contracts for difference’ they can be agreed to with either multi- or single-period
settlement terms.
Bulk Commodities
363
Voyage charter (VC)
Effective date:
To start 3.5 years after trade date
Maturity:
12 months (1 January–31 December) from effective date
Calculation period:
Monthly
Settlement:
Monthly
Notional quantity per period: 10,000 MT
Total notional quantity:
120,000 MT
Fixed price payer:
Client
Contract rate:
USD 11.00/MT
Fixed amount:
Contract rate x notional quantity per period
Floating price payer:
Bank
Floating amount:
Floating price x notional quantity per calculation
period
Floating price:
Baltic C4, average of daily published rates expressed in
USD/MT
Key points:
▪ The transaction is forward starting in that it only becomes effective 3.5 years in the
future.
▪ It is a single transaction covering a 12-month period with monthly cash flow
exchanges.
▪ The notional amount is expressed in metric tonnes.
▪ The fixed monthly cash flow payable by the client is USD 11.00 x 10,000 MT = USD
110,000. Sometimes the entity paying fixed is referred to as the buyer.
▪ Although not specified in this term sheet, the client may be using the swap to hedge
an underlying fixed price transaction. Suppose they have a physical transaction
where they have agreed to a fixed price voyage charter with a mining company.
Under the terms of the underlying contract they would receive a fixed cash flow
from their counterparty, but could then make a fixed payment to the bank under the
terms of the swap. In return they would receive a variable cash flow, which would
be beneficial if they thought that freight rates were going to be higher during the
period of the swap.
▪ The floating price is referencing the Baltic C4 route (Richards Bay to Rotterdam),
which forms part of the Capesize 2014 index. The transaction requires a price rather
than an index value. The calculation of an index or monetary value was illustrated
in Section 10.6.4.
▪ The value for the Baltic C4 route is published on every good business day of the
month (typically about 21 days), and to derive the settlement value an unweighted
arithmetical average of all these values is then calculated.
▪ Suppose that the floating price for a particular month fixes at USD 11.50/MT. The
floating cash flow will be USD 11.50 x 10,000 MT = USD 115,000. The payer of
floating is sometimes referred to as the seller.
▪ Since both payments coincide there will be a single net payment by the payer of
floating (in this case the bank) of USD 5,000.
364
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Time charter (TC)
Effective date:
Maturity:
Calculation period:
Settlement:
Total notional quantity:
Notional quantity per period:
Multiplier:
Fixed price payer:
Contract rate:
Fixed amount:
Floating price payer:
Floating price:
Floating amount:
To start 1.25 years after trade date
12 months (1 January–31 December) from effective
date
Monthly
Monthly
182.50 calendar days
The number of days in the calculation period.
0.5
Bank
USD 15,000 per calendar day
Contract rate x (notional quantity per calculation
period x multiplier)
Client
Baltic Supramax S2, average of daily published
prices expressed in USD/day
Floating price x (notional quantity per calculation
period x multiplier)
Key points
▪ Like the VC swap, this transaction is also forward starting with monthly settlement
covering a 12-month period.
▪ The notional amount is described in days and this example covers half of the transaction’s maturity.
▪ The fixed amount payable in a month with 31 days would be USD 15,000 x (31 x
0.5) = USD 232,500.
▪ Suppose that the floating price fixes at USD 14,500. The floating amount payable
for the same 31-day month would be USD 14,500 x (31 x 0.5) = USD 224,750.
▪ Since both payments coincide there will be a single net payment by the payer of
fixed equal to USD 7,750.
▪ This transaction includes the use of a multiplier, which reduces the number of days
within the month to which the swap applies and so lowers the cash flow. The use
of the multiplier could suggest that the counterparty assumes that the vessel(s) will
only be chartered for half of the time.
▪ Not all TC transactions will use a fractional multiplier and so in some cases it is set
at a value of 1.
▪ Although not specified by the term sheet, it may be possible to infer the client’s
motivation based on some simple assumptions. If the client were chartering their
vessels to other market participants, the fact that they are ‘selling’ the swap would
suggest that they have a concern that rates are likely to fall. It could be a pre-emptive
hedge in that the client knows they will have an ongoing need to charter vessels to
other participants and so this may not be linked to a specific transaction.
365
Bulk Commodities
Worldscale
Effective date:
Maturity:
Calculation period:
Settlement monthly:
Total notional quantity:
Notional quantity per period:
Fixed price payer:
Contract rate:
Fixed price:
Floating price payer:
Floating reference price:
Floating amount:
To start one month after trade date
Three months (1 January–31 March)
Monthly
Monthly
15,000 MT
5,000 MT
Bank
178 Worldscale points
A price calculated using the following formula
USD 24.7954/MT (being the contract rate of 178 x
Worldscale flat rate/100)
Client
Baltic Dirty Tanker Index route TD8 (Kuwait to
Singapore), expressed in Worldscale points.
Average of daily prices in each calculation period.
The unweighted arithmetic average of the reference
price for each good business day in the month x
Worldscale flat rate/100.
Key points
▪ Since the confirmation states that the fixed contract rate is WS178 and that this
converts into a cash flow of USD 24.7954 it follows that the ‘flat’ Worldscale rate
(e.g. WS100) for this route is USD 13.93/MT.
▪ The value of the Baltic Dirty Tanker Index is assessed from 14 individual routes (of
which TD8 is one), where the main cargo is typically crude oil. Each route and the
overall index can be quoted either in Worldscale points or as an index equivalent.
▪ The fixed price payment for each month will be the same and is calculated as USD
24.7954 x 5,000 MT = USD 123,977.
▪ Each good business day (usually about 21 days in a calendar month) a Worldscale
value for the route will be assessed and published. At the end of the month, the
values are then totaled and divided by the number of good business days to generate
the arithmetic average.
▪ Suppose that the floating reference price for the month is calculated at WS170. The
floating amount would be USD 13.93 x 1.70 x 5,000 MT = USD 118,405.
▪ Since both payments coincide there will be a single net payment by the payer of
fixed equal to USD 5,572.
FFAs can offer a great deal of flexibility:
▪ The participants can choose the size of the trade and it does not have to exactly
match the underlying exposure.
366
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ The maturity can be agreed between the two counterparties and again does not
have to match the underlying.
▪ A position can be added to or reduced. Reducing a position may mean taking an
offsetting position.
Although the emphasis in this section has largely been on hedging applications, it
is possible to use FFAs to express views on how the market may evolve without holding
an underlying exposure. Example trades may be:
▪ The spread in price between different maturities of the same contract; sometimes
referred to as the ‘calendar spread’. So, in the early part of a particular year a trader
could take a view that prices in the third quarter may move relative to the fourth
quarter. The periods traded could also include half and full years.
▪ Spreads between indices (e.g. Capesize 2014 vs. Panamax).
▪ Spreads between wet and dry indices (e.g. Baltic Dry Index vs. Baltic Dirty Tanker
Index).
10.6.7.2
Liquefied natural gas (LNG)
LNG was covered in Chapter 7 and one of the highlighted themes was the move away
from long-term contracts priced against oil to spot transactions priced using the gas
markets. In the long-term contract model the cost of freight was usually included as
part of the overall terms. As of the time of writing it has been suggested that around
20% of all LNG trades now take place on the spot markets (Risk, 2020), which usually
means that freight costs will be a separate variable cost.
An LNG trader could buy an LNG freight future so they could buy the natural gas
from a particular region where it may be relatively inexpensive and then later charter
a vessel in the spot market. The long freight future position would result in a known
shipping cost.
Freight futures on the NYMEX exchange reference three main routes:
▪ Gladstone, Australia to Tokyo, Japan;
▪ Sabine, USA to Isle of Grain, UK;
▪ Sabine, USA to Tokyo, Japan.
The contracts are listed out to two years and are financially settled against a reference index published by the Baltic Exchange.
10.6.7.3
Exotic swap structures
Target payout structures have been popular in a variety of different asset classes such
as foreign exchange and have been seen within the freight market, although to a much
lesser extent.
367
Bulk Commodities
A hypothetical term sheet might look as follows:
Freight Index:
Put price payer:
Call price payer:
Strike:
Maturity:
Calculation period:
Settlement:
Notional quantity:
Put payment:
Call payment:
Target payout provision:
Target payout amount:
Floating price:
Panamax Time Charter 4 (average of P1A_03, P2A_03,
P3A_03, P4_03)
Client
Bank
USD 10,000 per calendar day
Two years – subject to target payout provision
Monthly
Monthly
Number of days in the calendar month
Notional quantity x MAX [0, Strike – Floating price] x 2
Notional quantity x MAX [0, Floating price – strike],
subject to target payout provision
If the amount paid out under the call payments is equal
to or exceeds the target payout amount then the trade
will terminate.
USD 350,000
The average realised freight index value during the
month.
Looking at the structure, the term sheet indicates that it comprises of a strip of
options. Each month will consist of a combination of calls and puts. The client buys a
call and finances its purchase with the sale of a put. Since the call and the put have the
same strike, the client is therefore long a synthetic forward. Note that the client has sold
twice the amount of put optionality because the strike of the structure is set at a level
that is more favourable than the vanilla rate. This advantage is paid for by a combination
of a limit on the customer’s upside benefit and an increase in their downside liability.
This trade was designed with a client who would need to charter a vessel and is
looking to manage the associated cost. In essence the client is long a forward, which
will terminate once the accumulated positive payouts to them has reached the target
profit amount. The client cannot receive more than the target profit amount. If there is
a decline in prices the client will be faced with having to settle a liability, which will be
the difference between the now lower price and the strike price. However, this payment
is magnified by a factor of two. Although there is a risk that the cash flow paid under
the swap will increase, there should be a fall in the associated ‘physical’ market cost.
10.6.7.4
Option structures
Option structures have been traded in the freight market. Example structures would
include single option positions (e.g. calls and puts) as well as premium reduction strategies (e.g. zero premium collars, three ways). Rather than repeat ideas covered earlier in
the text, readers are referred to Chapter 5 on base metals for example trades.
CHAPTER
11
Climate and Weather
11.1
THE SCIENCE OF CLIMATE CHANGE
11.1.1
Definitions
In the first edition of this text, the author used the phrase ‘global warming’, which at
the time was a commonly used term. However, over time this has been superseded by
the phrase ‘climate change’.
I often feel there is a tendency for people to confuse weather and climate, so it is
worth defining both terms. NASA’s (2020) global climate change website provides some
useful terms of reference:
‘Weather refers to atmospheric conditions that occur locally over short periods of
time; from minutes to hours or days. Familiar examples include snow, clouds, winds,
floods, or thunderstorms. Climate, on the other hand, refers to the long term regional
or even average temperature, humidity and rainfall patterns over seasons, years, or
decades. Global warming is the long-term heating of Earth’s climate system observed
since the pre-industrial period (between 1850 and 1900) due to human activities,
primarily fossil fuel burning, which increases heat-trapping greenhouse gas levels in
Earth’s atmosphere. The term is frequently used interchangeably with the term climate
change, though the latter refers to both human- and naturally-produced warming and
the effects it has on our planet. It is most commonly measured as the average increase
in Earth’s global surface temperature.’
11.1.2
Greenhouse Gases
The earth is warmed by energy radiated by the sun. Although some of this is reflected
back into space, a significant proportion will reach the surface of the earth. Most of the
energy is absorbed by the earth’s surface but some is reflected back into space in the
form of infrared radiation. This infrared radiation, however, may not directly exit into
space, as its eventual escape may be blocked by greenhouse gases (GHG) that make up
about 1% of the atmosphere. The main greenhouse gases are:
▪ Water vapour (H2 O) – the most abundant greenhouse gas.
▪ Carbon dioxide (CO2 ) – this is generated by burning fossil fuels and by the respiration of humans and animals.
368
369
Climate and Weather
▪ Methane (CH4 ) – this may be released when mining for coal or searching for natural
gas and oil. It can also be generated by landfills and livestock.
▪ Nitrous oxide (N2 O) – this can be generated by burning fossil fuels or may be found
in fertilisers.
▪ Hydrofluorocarbons (HFCs) and perfluorocarbons (PFCs) – produced by refrigeration and air conditioning units.
▪ Sulfur hexafluoride (SF6 ) – which is used in a variety of manufacturing processes.
The focus of the climate change debate to date has been on carbon dioxide emissions. As a result, it is common to express the amount of these other GHGs in terms of
tonnes of CO2 equivalent. (CO2 e). This is a measure for comparing the radiative effect of
different GHGs in terms of the corresponding impact of emitting carbon dioxide. Carbon dioxide equivalents are calculated by multiplying the Global Warming Potential
(GWP) of a gas by its emitted weight.
Name of gas
GWP
Carbon dioxide
Methane
Nitrous Oxide
Hydroflourocarbons
Perflourocarbons
Sulfur hexafluoride
1
25
298
Up to 14,800
Up to 12,200
22,800
GHGs are not inherently bad as they trap heat and maintain the planet some 30 ∘ C
warmer than it would otherwise be. Most greenhouse gases do occur naturally, but their
volume is increasing because of human activity. According to the IPCC (2014) the primary drivers of GHGs are:
Energy industry:
Agriculture, forests, and other land uses:
Industry:
Transport:
Building sector:
11.1.3
35%
24%
21%
14%
6.4%
The carbon cycle
The carbon cycle describes the possible pathways a carbon atom takes through the different components of the ecosystem. A convenient starting point is the atmospheric
content of carbon dioxide. Carbon atoms may be retained in the air until eventually
they will be absorbed by green plants, combined with water, and turned into starch.
This could occur over a matter of hours or hundreds of years.
370
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The carbon atom is now part of a large starch molecule. There are two possible
pathways the atom may now take:
▪ It may be used by the plant in respiration, thus releasing the carbon atom back into
the atmosphere as carbon dioxide.
▪ It may be further processed and become part of the plant in the form of wood or
other plant cells or molecules.
If the carbon becomes part of the plant, it then may take one of three possible
pathways:
▪ The plant may die and be decomposed by bacteria in the soil, thus releasing the
carbon as carbon dioxide back into the atmosphere.
▪ An animal may eat the plant.
▪ If the plant is burnt (either directly or as a fossil fuel), carbon dioxide is released
back into the atmosphere.
If eaten by an animal (as part of a larger organic molecule such as fat, carbohydrate, or protein) it will then follow one of three pathways:
▪ The carbon is used in respiration and released into the atmosphere as carbon
dioxide.
▪ The carbon becomes part of the animal having been used to build protein, fat, or
carbohydrate.
▪ The carbon remains in the state it was eaten, is undigested, and is passed out of the
animal as faeces, which will be broken down by micro-organisms and the carbon
in them released once more as carbon dioxide into the atmosphere.
From here we start to ‘recycle’ in that the animal may die or be eaten by another animal and again the carbon atom can follow several different pathways. The carbon will
always eventually end up back in the atmosphere as carbon dioxide, which is why the
system is called a cycle, even though it does not really follow a single, simple pathway.
11.1.4
Feedback loops
The science of global warming is complex in that the effect of rising temperatures may
well be accompanied by other changes in climate. These changes, which are referred to
as feedback loops include:
▪ The amount of cloud cover.
▪ Levels of precipitation.
▪ Wind patterns.
▪ Rising sea levels.
▪ An intensification of the water cycle increasing the risk of droughts and floods.
▪ A change in the duration of the seasons.
Climate and Weather
371
Although some of these effects may give rise to cooling effects, such as increased
cloud cover that may block out sunshine; others may exacerbate the rise in temperatures. For example, if the Siberian permafrost were to melt exposing the peat surface
beneath, there would be an increase in the amount of methane released into the
atmosphere.
Man-made aerosols, which are microscopic particles that reflect sunlight back out
into space, may offer some temporary cooling effect but have their own side effects such
as a reduction in air quality.
11.2
THE CONSEQUENCES OF CLIMATE CHANGE
The debate over climate change has evolved from whether the earth is getting warmer to
one which tries to identify to what extent human activity has contributed to the change,
so-called anthropogenic activity. NASA (2020) provides a useful illustration of the balance that exists between an insufficient and excessive greenhouse effect. Mars has a
very weak greenhouse effect as it has a thin atmosphere, comprising mostly of carbon
dioxide. There is little methane or water vapour that would help reinforce the greenhouse and as a result the planet is largely frozen. Venus is an example of a planet with
a significant greenhouse effect. Like Mars its atmosphere is nearly all carbon dioxide,
but it has 154,000 times more of the gas as Earth. This produces ‘a runaway greenhouse
effect and a surface temperature hot enough to melt lead’.
NASA argues ‘The consequences of changing the natural atmospheric greenhouse
are difficult to predict but certain effects seem likely’.
▪ On average the Earth will become warmer. Although warmer temperatures may
benefit some regions, for others it will have a negative impact.
▪ Warmer conditions will probably lead to more evaporation and precipitation overall, but this will vary between regions. Some will become wetter, others dryer.
▪ A stronger greenhouse effect will warm the oceans and partially melt glaciers and
other ice increasing the sea level. Ocean waters also expand if it warms, contributing further to sea level rises.
▪ Some crops and other plants may respond favourably to increased CO2 as they will
grow more vigourously and use water more efficiently. However, changing climate
patterns may alter the areas where crops grow best.
11.2.1
Fifth assessment report of the IPCC
The fifth assessment report of the Intergovernmental Panel on Climate Change
(IPCC) was published in 2014. The main findings of the report included the following
observations:
▪ Human influence on the climate system is clear, and recent anthropogenic emissions of greenhouse gases are the highest in history. Recent climate changes have
had widespread impacts on human and natural systems.
372
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Continued emission of greenhouse gases will cause further warming and
long-lasting changes in all components of the climate system, increasing the
likelihood of severe, pervasive, and irreversible impacts for people and ecosystems.
Limiting climate change would require substantial and sustained reductions in
greenhouse gas emissions which, together with adaptation, can limit climate
change risks.
▪ Adaptation and mitigation are complementary strategies for reducing and managing the risks of climate change. Substantial emissions reductions over the next few
decades can reduce climate risks in the twenty-first century and beyond, increase
prospects for effective adaptation, reduce the costs and challenges of mitigation
in the longer term and contribute to climate-resilient pathways for sustainable
development.
▪ Many adaptation and mitigation options can help address climate change, but no
single option is sufficient by itself. Effective implementation depends on policies
and cooperation at all scales and can be enhanced through integrated responses
that link adaptation and mitigation with other societal objectives.
11.3
THE ARGUMENT AGAINST CLIMATE CHANGE
‘Nobody knows just how much carbon dioxide the world is going to produce
in the future. Nobody knows just what it will do to the temperature. Nobody
knows just how temperature rises will affect the world economy’.
—The Economist – 4 November 2006
Although there has been a growing consensus that human activities have contributed significantly to the warming of the earth, there is still an undercurrent of disagreement on the subject. The main reasons for disagreement are as follows:
▪ Climate change has been defined solely in terms of the study of the effect of greenhouse gases. One argument is that solar activity is a significant contributor to the
climate and that there is a possibility that a reduction in the sun’s activity could
lead to significant cooling. This argument is disputed by bodies such as NASA.
▪ Just because most people believe in something, this does not necessarily mean that
it is right.
▪ The main research that supports the current majority theory is not independent
and is being funded by those with an interest in propagating the current consensus
view.
▪ Obtaining funding for scientific research that may contradict global warming is difficult to procure.
▪ As a result of the current concern over climate change, a new industry has developed, which many individuals and institutions have an interest in perpetuating.
▪ Any information that may contradict current thinking is not given the prominence
that it deserves. For example, those arguing against the extent of anthropogenic
activity point out that:
▪ The earth underwent a similar period of warming from 1918 and 1940.
Climate and Weather
373
▪ The earth cooled from the 1940s to the 1970s. This is countered, however, by the
argument that this was caused by the cooling effect of sulfur in the atmosphere
generated by industrial activity, which is now in decline.
▪ Crops in areas normally associated with warm temperatures have been ruined
by unseasonably cold weather.
▪ Some areas in East Antarctica are becoming colder.
▪ Between 1998 and 2005, global average temperatures did not increase.
▪ The choice of language in the predictions is too vague: ‘Since the early 1990s, the
columns of many leading newspapers and magazines, worldwide, have carried an
increasing stream of alarmist letters and articles on hypothetical human-caused climate change. Each alarmist article is larded with words such as ‘if’, ‘might’, ‘could’,
‘probably’, ‘perhaps’, ‘expected’, ‘projected’, or ‘modelled’.
▪ Even with a high probability of an event occurring, the margin for error is still significant, ‘a 10% uncertainty in any theory is a wide open breach for any latter-day
Galileo or Einstein to storm through with a better idea’.
▪ The science is too complex to reconstruct past global temperatures and model them
to predict future outcomes. The contrary argument says that the climate will change
naturally, sometimes predictably sometimes unpredictably. The changes could be
gradual, while others will be shorter and difficult to explain. The IPCC has been
criticised for using 30 different computer models none of which produce the same
result (Financial Times, 2020).
▪ The time frames over which climate judgments are made are often selective. For
example, over 90% of the last two million years, the climate has been much colder
than it has been today and if looked at from this perspective the planet is emerging
from an ice age.
▪ There is a lack of information from the Southern Hemisphere.
▪ A report on the economics of climate change issued by the UK House of Lords
Select Committee on Economic Affairs (prior to the publication of the 4th Intergovernmental Panel on Climate Change report) included the following comments:
▪ There was concern about the objectivity of the IPCC process in that some of the
scenario analysis and summary documentation was influenced by political considerations.
▪ Some of the positive aspects of global warming were downplayed in the third
IPCC report.
▪ The preoccupation of setting emissions targets at an international level was not
an effective method of addressing the climate control issues. ‘The Kyoto Protocol
makes little difference to rates of warming and has a naïve compliance mechanism which can only deter countries from signing up to subsequent tighter
emission targets’.
11.4
HISTORY OF HUMAN ACTION AGAINST CLIMATE CHANGE
11.4.1
Formation of the IPCC
The first world conference on climate change took place in 1979, but it was not until
1988 that the World Metrological Organisation and the United Nations Environment
374
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Programme established the IPCC. The IPCC reviews published research on the issue of
climate change and issues influential assessments on a periodic basis and to date they
have issued reports in 1990, 1995, 2001, 2007, and 2014.
11.4.2
The Earth Summit
The IPCC’s first report in 1990 formed the basis of negotiations for the United Nations
Framework Convention on Climate Change. The text of the convention was launched
in June 1992 at the Rio de Janeiro ‘Earth Summit’. The convention identified a number
of countries referred to as Annex I parties, which comprised the industrialised nations
that were members of the OECD in 1992 plus countries that were deemed to be in
transition (e.g. the Russian Federation). This listing would eventually determine how
many countries needed to ratify the subsequent Kyoto Protocol for this to become legally
enforceable.
11.4.3
The Kyoto Protocol
The second IPCC assessment report concluded, on the balance of available evidence,
that human activity was influencing the climate and that this would pose a threat to
human and economic development. These results paved the way for the development
and signing of the Kyoto Protocol in December 1997. The Protocol set individual, legally
binding targets for several industrialised countries (the Annex I countries) willing to
take steps to curb their emissions of greenhouse gases.
For the Protocol to be legally binding it had to be ratified by an agreed number of
developed countries who were determined to be responsible for most global emissions
in the industrialised world. Only those countries that ratified the protocol become parties to the treaty, but ratifying the Protocol did not require the signatory to adopt legally
binding emissions targets unless they were an Annex I country.
The Kyoto Protocol outlined a variety of mechanisms that would allow countries to
meet their commitments. Each of the countries for which the Kyoto principle is legally
binding were assigned an emission limitation or reduction target. The emission targets varied from country to country and ranged from a cut of 8% to an increase in 10%
of a baseline value measured from 1990. Despite the different target ranges, the main
aim of the accord was to achieve an overall reduction in greenhouse gases by 5% in the
2008–2012 period, termed the ‘Kyoto period’.
The Protocol outlined three mechanisms to help achieve this aim. The three Kyoto
mechanisms were:
▪ Emission Trading Schemes (ETS)
▪ Clean Development Mechanism (CDM)
▪ Joint Implementation (JI)
Two of the mechanisms (CDM and JI) are project-based, while the third is a
market-based mechanism (ETS). These mechanisms, however, were never intended
as the only way in which countries would tackle emissions. In addition, countries are
Climate and Weather
375
expected to pursue policies that would encourage an overall reduction in the levels of
emissions.
Emission trading schemes – The market for trading emissions involves the purchase
or sale of the right to emit an agreed amount of a specified pollutant. Emission trading
schemes are sometimes referred to as ‘carbon markets’ as carbon dioxide is the most
widely produced greenhouse gas. Participants can use the mechanism to acquire rights
in order to meet their emission targets.
Under the terms of the Kyoto Protocol these rights were referred to as ‘assigned
amount units’ (AAU) where one AAU is equal to one tonne CO2 e. Hence if an installation had emitted 100 tonnes CO2 e in 1990 then over the Kyoto period of 2008–2012 they
would be faced with a target of, say 92 AAU. If the installation chose to emit more than
this amount, they would either have to buy the necessary allowances in the market or
invest in technology to reduce their level of emissions. Within the EU Emissions Trading Scheme (ETS), these allowances are termed EUAs (European Union Allowances)
and are either allocated or auctioned by regulators.
The EU ETS is an example of what is known as a ‘cap and trade’ scheme. A national
regulator allocates allowances to different sectors and then individual eligible firms. The
regulator will be responsible for monitoring compliance within the limits and will also
enforce penalties for non-compliance.
Clean development mechanism (CDM) – this represents investment by an Annex
I country in a developing country that will result in measurable long-term climate
change. CDM projects, which became effective in 2005, must reduce emissions below
those emissions that would have occurred in the absence of the project. These projects
are monitored and certified by the UN, who will then issue credits on an annual basis
equal to the amount by which emissions have been reduced.
CDM credits are referred to as certified emission reductions (CERs). The CERs generated by such project activities could be used by eligible entities in Annex I countries to
help meet their emissions targets under the Kyoto Protocol or be traded on a secondary
market basis.
Typical projects include:
▪ Energy efficiency schemes.
▪ Methane capture.
▪ Fuel switching (e.g. from coal to gas).
▪ Capture and destruction of certain industrial gases.
According to the United Nations Framework Convention on Climate Change
(UNFCCC) 2018 annual report the CDM involved 140 countries with 7,803 projects.
Joint implementation (JI) – In this mechanism, one Annex I country invests in
an emission reduction project in another Annex I country. The costs of cutting emissions in the target country is cheaper than in the investing country. This allows for
the transference of allowances from the target country to the investing country. These
allowances are referred to emission reduction units (ERUs), but unlike the clean development mechanism there is no overall change in emissions. The JI mechanism took
effect from 2008 to coincide with the Kyoto period.
376
11.4.4
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
From Kyoto to Paris
After Kyoto, the next significant landmark agreement was at the ‘Conference of Parties’
(COP) 21 in Paris in 2015. The agreement came into force in November 2016 and as at
the time of writing had been ratified by 96% of the parties to the convention. Some of
the key elements of the agreement were:
▪ A long-term temperature goal – The aim is to limit global temperature increases to
well below 2∘ C, while pursuing efforts to limit the increase to 1.5∘ C.
▪ Global peaking and climate neutrality – To achieve the temperature goal, parties to
the agreement aim to reach a global peak of GHGs as soon as possible.
▪ Mitigation – The Paris Agreement establishes binding commitments by all parties to
prepare, communicate, and maintain a nationally determined contribution (NDC)
and to pursue domestic measures to achieve them.
▪ Sinks and reservoirs – parties are encouraged to conserve and enhance GHGs sinks
and reservoirs.
▪ Finance and technology – The Paris Agreement reaffirms the obligations of developed countries to support the efforts of developing country parties to build clean,
climate-resilient futures.
▪ Global stock take – To take place in 2023 and every five years thereafter, will assess
collective progress toward achieving the purpose of the Agreement in a comprehensive and facilitative manner.
▪ Mitigation mechanisms – The accord aims to establish a new mechanism that will
replace the CDM and JI frameworks.
According to the Intergovernmental Panel on Climate Change, in order to limit
warming to 1.5∘ C, CO2 emissions will have to fall by about 45% by 2030, relative to
their 2010 levels. Limiting global warming to 2∘ C will require a transition to become a
carbon-neutral economy by the middle of this century.
11.5
PRICE DRIVERS OF EMISSIONS MARKETS
Like all commodities the market clearing price will be determined by the interaction of
demand and supply (Figure 11.1). In its simplest form, market participants will demand
more of a product as the price falls. On the supply side, participants will provide greater
volume in response to higher prices. The equilibrium or market clearing price is where
demand and supply interact.
However, within emissions markets the interaction of these two variables is slightly
different than the classic economic representation. Figure 11.2 illustrates a possible
interaction.
The supply curve for emission allowances will generally be fixed on an annual
basis although it is possible that certain schemes may allow for the use of ‘offset credits’
(e.g. CERs). The demand curve initially follows the traditional representation of sloping from top left to bottom right, but there are one or two anomalies. If demand exceeds
supply, then participants will usually face a fixed fine, say, EUR 100.00. The level of
377
Climate and Weather
Supply
Price
Price
Demand
Volume
FIGURE 11.1 Classic demand and supply curves.
Fixed supply
of allowances
Price of
carbon
Price
Demand for
allowances
Volume of
credits
FIGURE 11.2 Demand and supply within the emissions markets.
Source: European Commission.
this fine would place an upper limit on the price of allowances. In this scenario, no one
would pay, say, EUR 110.00 for an allowance if the fine is set at a level that is EUR 10.00
lower. Also note the demand curve becomes more smoothed at higher prices. This
could be reflective of the fact that allowances will always have some monetary value and
if not used in the current compliance period could be carried over to the next period.
Economic growth – As an economy expands it is likely that production of carbon and
other industrial gases will increase. For action on global warming to be truly effective,
an overall global limit on the amount of greenhouse gases produced would be needed.
Politically this would be exceptionally difficult to achieve, as emerging countries such
as India and China are unlikely to agree.
378
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Supply of allowances – If the supply of allowances is significantly larger than actual
emissions then the price will fall. The aim of such schemes is to make sure that the
supply is lower than the verified emissions to encourage a reduction in emissions perhaps by the implementation of new technology. The supply of allowances may also be
impacted by political rather than economic decisions.
Demand for allowances – The choice faced by installations subject to emission targets is whether they should invest in technology to reduce their emissions, therefore
being able to sell any surplus allowances, or to simply buy their allowances from another
market participant. Logically, participants will choose the lower cost option of either
investing in abatement technology or buying allowances.
Weather patterns and fuel prices – Since one of the biggest sources of emissions
is power generation, weather patterns and fuel prices will influence the price of
allowances. If a generator has the ability to substitute different fuel sources, such as
coal or natural gas, they may compare the carbon-adjusted dark and spark spreads
to determine the optimum fuel source. If the price of natural gas were to increase,
combined with a period of cold weather, a generator may be encouraged to switch
to coal as the primary fuel input. Since coal generates more greenhouse gases, the
generators would need to buy more allowances to cover the increased use of the
commodity. This should then drive up the price of allowances, the cost of which would
be passed onto consumers in terms of a higher electricity price.
Intuitively, there is a positive correlation between the prices of carbon and the prices
of both power and gas. There is a negative correlation between the price of coal and the
price of carbon.
Penalties for non-compliance – These are determined by each country or region.
Fees for non-compliance in the EU scheme have been established but irrespective of
the charge, any breach of the targets will need to be made up in the following year.
Alternative sources of electricity production – The cost of emissions will be affected
by the planned increase in alternative low carbon technologies such as nuclear power
and renewable sources. If such technological changes are expensive this will lead to an
increase in the demand for allowances.
Regulation – Since this is a business that is driven by government regulation there
is always the risk that politicians may choose to change the rules of the game. Factors
that might affect the price of allowances could include:
▪ The stringency of allowance allocations.
▪ Whether project-based credits can be included and the extent to which they could
be used.
▪ Whether allowances or credits can be carried over between compliance periods.
▪ The perceived strength of the compliance regime used to police a particular scheme.
Linkages with other energy markets – In an article entitled ‘The Energy Revolution’, Bond (Barclays, 2007) provides an interesting insight into the linkages that exist
between different energy markets and the price of carbon. He argues that the early
part of the twenty-first century may ‘be seen as an inflexion point in the global economic
order’. He points out that the market is signaling that the supply of conventional energy
Climate and Weather
379
sources needs to be increased, but political sentiment is moving away from the reliance
on hydrocarbon-based fuels.
‘Put bluntly the world has simultaneously discovered that the supply of energy
may not match future demand and that the type of energy currently supplied
is incompatible with survival. Market prices signal the need for dramatic
increases in the supply of hydrocarbon-based energy, while political shifts signal the need for an equally dramatic decrease in the use of hydrocarbon-based
energy’.
—Barclays, 2007
Furthermore, he argues the global economy needs to invest heavily in the energy
infrastructure to meet a projected 50% increase in demand over the next 30 years. To
avoid the worst consequences of global warming, however, there needs to be an 80% cut
in energy supplied from fossil fuels to alternative sources.
Bond points out that a positively sloped forward curve (i.e. a market in contango)
signals a longer-term excess of demand over supply; a curve in backwardation would
therefore be indicative of a shorter-term imbalance between markets (demand exceeding supply) and a longer-term excess of supply over demand. At the time the piece was
written, forward curves for the energy markets that emit the most CO2 were in contango
(e.g. coal, crude oil) while the curves for cleaner energy were backwardated (e.g. natural
gas). Based on these observations he concluded:
‘We can deduce that as yet, the markets do not believe that policy changes
will successfully reduce global demand for the dirtier fuels in favour of cleaner
energy sources’.
Although the quote may seem somewhat dated, there is a wider point in that interpreting forward curves in this manner can give some insight into the markets’ collective
view on the efficacy of action on climate change.
11.6
EU EMISSION TRADING SYSTEM
‘The carbon market works like any other commodity market; companies trade
and the market sets prices. But it is unusual in that the commodity being
bought and sold does not exist: it is the certified absence of carbon emissions’.
—The Economist – 9 September 2006
11.6.1
Background
The EU Emissions Trading System (ETS) was launched in 2005 and was the first ever
emissions market. The market has evolved through different stages:
▪ Phase 1 (2005–2007)
▪ Phase 2 (2008–2012)
380
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ Phase 3 (2013–2020)
▪ Phase 4 (2021–2030)
The system works on a cap and trade principle. This means that a cap is set on
the total amount of GHGs that can be emitted by installations who operate within
the scheme. This cap is designed to fall over time to achieve an overall reduction in
emissions. The other main feature is that participants are either allocated or must buy
allowances that permit to emit GHGs. Any unused allowances can be traded within the
marketplace.
Worked example:
Let us say that two manufacturing companies A and B both emit 100,000 tonnes
of CO2 per year and each entity is given 95,000 emission allowances. One allowance
represents the right to emit one tonne of CO2 . So, neither company is fully covered for
its emissions. At the end of each year, the companies must surrender the number of
allowances that correspond the amount they have emitted during the year. If they fail
to do so, they face a fine of EUR 100 per tonne. Companies A and B do not want to pay
the fine and both must manage the excess 5,000 tonnes of CO2 . They have two ways of
doing this.
They can either reduce their emissions by 5,000 tonnes, or purchase 5,000
allowances in the market, either in the secondary market or via an auction. To decide
which option to pursue, they will compare the costs of reducing their emissions by
5,000 tonnes with the market price for allowances.
Suppose that the allowance market price is EUR 10.00 per tonne of CO2 . We
will also assume that Company A’s reduction costs are EUR 5.00 (i.e. lower than the
market price). Company A will reduce its emissions because it is cheaper than buying
allowances. Company A may even reduce its emissions by more than 5,000 tonnes, say,
10,000 tonnes. For Company B, the situation may be the opposite: its reduction costs
are EUR 15.00 (i.e. higher than the market price) so it will prefer to buy allowances
instead of reducing emissions.
Suppose Company A spends EUR 50,000 on reducing 10,000 tonnes at a cost of EUR
5.00 per tonne and receives EUR 50,000 from selling 5,000 excess allowances at the prevailing market price of EUR 10.00. So, from the perspective of Company A, the net cost
to reduce their emissions by 10,000 tonnes is zero. In the absence of a trading scheme,
they would have been forced to reduce their emissions by 5,000 tonnes and their net
cost would have been EUR 25,000. From Company B’s perspective the cost of reducing
their emissions by 5,000 tonnes would be EUR 50,000 (buying 5,000 allowances at a
price of EUR 10.00). In the absence of the trading scheme, their cost of reducing their
emissions would be EUR 75,000 (5,000 tonnes at EUR 15.00).
11.6.2
System design
At a conceptual level, the design of such a scheme is relatively straightforward:
▪ What emissions will be targeted?
▪ Which entities will be covered by the scheme?
▪ How is the right to emit GHGs evidenced?
Climate and Weather
381
▪ How will the allowances be distributed?
▪ How is ownership of allowances monitored?
▪ What is the level of the cap?
▪ How are emissions monitored and verified?
▪ How long will an allowance last if it is not used for compliance purposes?
▪ Should the system be linked to similar programs in different geographical
locations?
Which emissions will be targeted?
The system focuses on carbon dioxide (CO2 ), nitrous oxide (N2 O), and perfluorocarbons (PFCs).
Which entities will be covered by the scheme?
The sectors that are covered include oil refineries, steel works, iron, and steel works, aluminium producers, cement producers, glass, ceramics, pulp, paper, cardboard, acids,
and bulk organic chemicals. From 2012, airlines became subject to the rules of the
scheme. All airlines operating in Europe (European or otherwise) are required to monitor, report, and verify their emissions.
How is the right to emit GHGs evidenced?
Participants can only emit GHGs if they hold sufficient allowances. For the EU ETS
these are referred to as European Union Allowances (EUAs). If they emit more than
the allowances, they hold then they will be subject to a fine of EUR 100.00 per tonne.
Each allowance gives the holder the right to emit one tonne of carbon dioxide or the
equivalent amount of N2 O and PFCs. In the absence of global agreement as to how
airline emissions should be handled the EU ETS decided to issue a separate category
of allowances to airline operators referred to as European Union Aviation Allowances
(EUAA). Airlines may use EUAs to meet compliance requirements, but the opposite
does not hold, i.e. non-airline entities cannot use EUAAs to satisfy their emission commitments.
How will the allowances be distributed?
In phase 1, virtually all allowances were given away from free. This proportion fell to
about 90% in phase 2 as the preferred method of auctioning allowances started to take
effect. Although some allowances will still be given away for free the figure subject to
auction is expected to gradually increase. This is based on the principle that the polluter
should pay. Allowances are distributed on an annual basis and the main auction platforms are the European Energy Exchange (EEX) and ICE Futures Europe. The revenues
generated from the sale of the auctions are returned to each EU member state and can
be used to finance other climate-related projects. Typically, a proportion of the annual
cap is auctioned every week. Free allowances are allocated at the end of February for
the current compliance year.
382
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Allowances originating from the auction process are referred to as ‘the primary
market’. The structure of the system allows the participants to transfer the ownership
of these allowances on a price basis, which is referred to as the ‘secondary market’.
The scheme also operates a New Entrants Reserve that can provide allowances
for new projects aimed to mitigate the impact of climate change (e.g. carbon capture,
renewables).
The 2008 financial crisis had a very serious impact on the EU ETS as the fall in
industrial production reduced the demand for allowances. By 2013 there was a surplus of 2.1 billion allowances, which resulted in very low prices. The EU addressed
this in two ways. The first was to postpone the auctioning of 900 million allowances
until 2019/2020. This is referred to as ‘back loading’ and did not reduce the number of
allowances that were auctioned during phase 3, but merely changed the timing of their
release. This reduced the surplus to 1.78 billion allowances by 2015.
The second method was the establishment of a market stability reserve (MSR),
which acts according to a set of pre-defined rules. If there is an excess supply of
allowances the MSR can restrict the amount offered for sale at the annual auction.
Currently, if the number of allowances in circulation exceeds 833 million then the
next auction will be reduced by a figure of 12% of the amount available. The MSR can
return the allowances under certain market conditions such as high prices (because
of high demand) or low supply. So, when there are 400 million allowances or less in
circulation a fixed annual amount of 100 million will be added to the amount to be
auctioned.
How is ownership of allowances monitored?
The EU has established an online registry that records the ownership of allowances of
each participant. Each participant will hold an account at the registry and any transfer
within the secondary market will be recorded. The registry also records the verified
emissions of each participant as well those allowances surrendered for compliance
purposes.
All transactions between accounts in the registry are checked and authorised by the
European Union Transaction Log (EUTL).
What is the level of the cap?
Phase 1 of the ETS was designed as a trial period and during this time there was no
reliable emissions data available to make a judgment as to the true level of emissions.
As a result, caps were set based on estimates. These estimates proved to be too optimistic
and as a result the market was oversupplied with allowances.
Phase 2 saw the introduction of a single EU-wide cap which replaced the previous
system of individual national caps. From 2021 onwards the overall number of emission
allowances will decline at an annual rate of 2.2%.
At the end of each year each participant must surrender enough allowances to cover
all its emissions or face a fine. If the company has an excess of emissions at the end of
the year it can use these allowances to cover its future needs or sell them to another
company that is short of allowances.
Climate and Weather
383
How are emissions monitored and verified?
The ETS compliance cycle is an annual process. Each participant much monitor and
report their emissions. The data for a given calendar year must be independently verified by 31 March of the following year. Once verified, the participant must surrender
the equivalent number of allowances by the end of the following month, i.e. 30 April.
The allowances are then canceled and cannot be used again. If actual emissions exceed
the number of allowances surrendered the participant is subject to a fine of EUR 100.00
per tonne.
How long will an allowance last if it is not used for compliance purposes?
If a market participant has more allowances than are needed for compliance purposes,
then broadly speaking they will have two options. They can sell any excess allowances
on the secondary market or retain the allowances for the next compliance period. ‘Banking’ an allowance is the process of retaining unused allowances in one period to be used
in the future. Any unused allowances need not be used in just the following year; they
can be used in any future compliance period.
Related to the concept of ‘banking’ is that of ‘borrowing’; this process allows a participant to use allowances from future periods to meet the current compliance requirements. Note that for the EU ETS, market participants must submit the required number
of allowances for the previous compliance period by 30 April. However, the allocation
of any free allowances for the current period is made by the end of February. In effect,
this allows those participants who enjoy a free allocation to borrow from the current
year to be compliant with the previous year.
Should the system be linked to similar programs in different geographical locations?
There is a movement towards linking trading systems around the world to encourage
the development of a global market. As of the time of writing this has not yet been
formalised. However, the EU ETS has allowed the partial use of international credits
generated from the CDM and JI to be used by participants towards fulfilling part of their
compliance obligations. It seems likely that this practice will be discontinued with the
start of phase 4.
11.6.3
Cap and trade versus carbon taxes
One subject that is often debated is whether the cap and trade systems should be
replaced by a direct tax levied by the government. Advocates of a tax argue that if
the number of permits allocated is misjudged the price of carbon would increase or
decrease significantly. Such price volatility may deter people from investing in green
technology. They also point out that if the scheme operates by allocating a fixed amount
(rather by auction), there is no adjustment for the fact that during a recession companies produce less and therefore pollute less. In the early phases of the EU ETS permits
were given away for free but were tradable on the secondary market. This allowed
some market participants to enjoy a windfall gain by selling their ‘free’ allowances.
As well as raising revenue, a tax provides a clear price floor for carbon and hence
a minimum return for investors willing to innovate in new green projects. Under a cap
384
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
and trade scheme an invention that reduced the cost of cutting carbon emissions could
push down the price of permits, reducing the investors’ returns.
11.7
EMISSION DERIVATIVES
11.7.1
Introduction
In its annual ‘State and Trends of Carbon Pricing 2019’, the World Bank noted:
▪ Globally, there were 57 carbon pricing initiatives implemented or scheduled for
implementation.
▪ These initiatives covered 11 gigatonnes (Gt) C02 e, which covered about 20% of GHG
emissions.
▪ Prices within the different schemes varied from USD 1.00–127/tC02e. However,
51% of the emissions covered were priced at less than USD 10.00.
▪ The various schemes raised about USD 44 billion in carbon pricing revenues.
For a participant who believes they will need to buy allowances, the decision as to
when they should be bought during the year requires them to develop a view as to how
prices could evolve during the period. They could:
▪ Buy the allowance for spot value and hold for the period.
▪ Buy the allowance for forward delivery. The price would be fixed at the trade date
but paid at the contract’s maturity. If the maturity date for the forward matches the
anticipated holding period for the spot ‘buy and hold’ transaction, then the costs
should be identical. However, fair value pricing does not always hold in the market
and is discussed in a subsequent section.
▪ Enter some form of option-based contract. Depending on the structure this may
require the payment of a premium
11.7.2
Spot transactions
EUAs can be sold on a spot basis with delivery occurring one day after trade date. This
could be done on an over-the-counter basis or via an organised exchange such as ICE or
the EEX. Trades tend to be executed on a standardised number of allowances: 10,000,
25,000, 50,000, and 100,000 being typical.
Futures
EUA futures can be traded on exchanges such as EEX and ICE. The specification
for the ICE EUA future is:
Unit of trading:
Currency:
Contract series:
Expiration date:
Settlement:
One lot represents 1,000 EUAs
EUR
Quoted with monthly, quarterly, or annual settlement.
Last Monday of the contract month.
Physical by delivery and transfer of the EUAs from the seller’s
account to the buyer’s account at the European Union Registry.
Climate and Weather
385
Forwards
Contracts traded on an OTC forward basis would be structured with a delivery date
that would be agreed between the two participants. For example, transactions that have
been traded include:
▪ A fixed price forward transaction would have similar characteristics to that of the
future. It would be a physically settled purchase or sale of the EUA at a fixed price
on a fixed date.
▪ A physically settled forward purchase at a fixed price for delivery on a pre-agreed
date but where the seller has the right to deliver the allowance on an earlier date.
If the seller delivers, they will receive a present-valued cash flow.
▪ A floating price forward will still commit the participant to either buy or sell the
underlying allowance on a pre-agreed fixed date, but the price at which the trade
would executed is not determined until sometime closer to the settlement date.
Typically, the contract price would reference an average of some agreed index value
such as a futures price.
▪ A floating price forward where the buyer has the right to nominate the volume of
EUAs it wishes to price prior to the determination of the contract price.
11.7.3
Forwards – fair value pricing
Since the EUA is 100% fungible, the cost of an allowance for settlement between two
different years in the same compliance period can be accounted for using standard
forward pricing techniques (see Chapter 2). That is, a forward price can be calculated as
the spot price plus the cost of borrowing money to buy the allowance. As of the day of
writing (March 2020) spot EUAs were trading at EUR 18.40, while delivery for March
2021 was EUR 18.15. The analysis is somewhat complicated by the fact that 12-month
wholesale borrowing rates were −0.25%, i.e. you can be paid to borrow money. A market participant who decides to buy a spot contract for EUR 18.40 and borrow the same
amount to finance it would be required to repay about EUR 0.05 less than they borrowed
(EUR 18.40 × −0.25% × 365 / 360). This would suggest that the fair value of the March
2021 contract should be EUR 0.05 lower than the March 2020 equivalent, i.e. EUR
18.35. The quoted market price is some EUR 0.20 lower than this ‘fair value’. Since firms
receive their allowances from governments more than a year in advance before they
need them for compliance purposes they could sell them for spot value (EUR 18.40) and
buy them back under the terms of a forward or futures agreement at EUR 18.15. The
cash proceeds they would be holding for the period would be economically equivalent
to borrowing money at a negative rate of interest (approximately 1.34%).
11.7.4
Repurchase agreements
An extension of the physically settled spot and forward transactions is a process known
as ‘inventory monetisation’. Under the terms of this transaction a corporate can use their
inventory of EUAs to borrow money at an attractive rate. This is an instrument that has
been used in several different financial and commodity markets for many years.
386
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Under the terms of the deal there will be two separate transactions executed simultaneously as a single package. The first deal is a spot transaction where the corporate
sells their inventory of allowances to a market participant such as a trader at an investment bank. Under the terms of the second deal, the corporate agrees to buy back the
same volume of allowances for forward delivery. The prices at which the two transactions are executed are determined using time value of money principles. Suppose that
a corporate decides to use this mechanism to borrow money for a one-year period for a
holding of one million allowances. A trader at the investment bank calculates that the
12-month forward price of the contract is EUR 15.00 meaning that the amount that the
corporate will have to repay at the end of the period will be EUR 15,000,000. The amount
that the corporate can therefore borrow at the start of the period is simply the present
value of this sum. The trader sees that 12-month interbank borrowing rates are 5.00%
and so decides to apply a credit spread of 0.50% to this rate. The present value of the forward proceeds is therefore EUR 14,207,772, which is calculated as EUR 15,000,000/(1 +
(0.055 x 365/360)). The impact of the deal is that the corporate is effectively borrowing
at a rate of 12-month Euribor plus 0.50%. Some readers may reasonably query why the
bank did not simply use the current spot rate to determine the initial proceeds. Recall
that in some commodity markets, the spot rate could be less liquid than the forward,
with the latter being perceived as a more accurate measure of fair value.
This trade would be advantageous to the corporate if they normally borrow money
at a rate greater than 0.50% over 12-month Euribor. The profit to the bank depends
on the strength of their balance sheet. To be able to finance the loan to the corporate,
the bank will have to borrow in the money markets and if they can achieve this at say
Euribor without any spread, they will also be able to profit.
11.7.5
Swaps
Emission swaps are financial in nature in that they are cash settled. They are only effective for one day and will involve an exchange of cashflows fixed for floating. In some
respects, they are like single day contracts for difference that are popular in the equity
markets. An indicative set of terms is shown below:
Trade date:
Effective date:
Termination date:
Pricing date:
Delivery month:
Commodity:
Total notional quantity:
Fixed rate payer:
Fixed price:
Floating price payer:
Reference price:
Settlement:
March (current year)
1 December (Year following trade date)
1 December (Year following trade date)
1 November (Year following trade date)
1 December (Year following trade date)
EU Allowance
25,000
Company A
EUR14.00
Bank
LEBA – CARBON – INDEX on the designated pricing date
for the relevant delivery date.
Cash
387
Climate and Weather
A swap contract would be useful if a participant wanted price protection but did not
require physical delivery. Equally the contract would be useful for those participants
who wished to express a view on the movement of allowance prices. It is financially settled, as it does not require ownership of an allowance to be transferred. The participants
agree a cash settlement amount based on the difference between two prices.
11.7.6
Physical and cash-settled options
Underlying commodity:
Notional quantity:
Option style:
Option type:
Option seller:
Option buyer:
Reference price:
Delivery month:
Option expiration:
Strike price:
Premium:
EUA allowance
30,000 allowances
European
Call
Bank
Client
ICE EUA futures settlement price for contract delivery
date
December
Nine months; to coincide with expiration date of
December future (in this case the last Monday of the
delivery month)
EUR 20.00 (At-the-money forward)
EUR 2.39/allowance. Total premium EUR 71,700
Notes:
▪ The transaction is an option on an exchange traded future.
▪ If the option is designated to be cash-settled the payoff to the call buyer will be:
MAX (Underlying price – strike, 0).
▪ If the option is designated to be physically settled, then if the buyer chooses to
exercise, they will take delivery of the agreed number of allowances that would
be transferred into their emission allowance account.
▪ The ICE futures expire on the last Monday of the delivery month but given Christmas holidays and the contract specification this may be the earlier in the month.
▪ Implied volatilities for EUAs can be significant with values sometimes reaching
more than 60%. A figure of 35% was used to value this option.
11.7.7
‘View driven’ strategies
Although the EUAs can be used to meet compliance with the emission targets, traders
will also be looking for opportunities to execute more speculative transactions. Some
simple trading strategies might include the following:
Identifying a mispriced contract – Say for example, a trader believed that a futures
contract due to mature in one year’s time was not trading at its fair value and was considered expensive. The trader could buy a spot contract and sell the forward. The trader
388
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
could take delivery of the spot contract, finance the holding for a 12-month period, and
deliver the allowance to meet the commitment created by the sale of the forward transaction. Buying and holding the allowance would realise the fair value of the position,
which should be less than the proceeds received from the sold forward.
Term structure slope trades – It is also possible to trade the futures term structure.
This would be the purchase (sale) of a short-dated contract combined with the sale (purchase) of a longer-dated contract. This is referred to as ‘trading the calendar spread’ and
is used to express a view on the price spread between two maturities. The trade would
be closed out when the trader believed that prices had come back into line.
Geographic arbitrage – If the price of an EUA is different on different exchanges in
different geographical locations (e.g. ICE vs. EEX) a trader could buy and sell to exploit
the differential.
EUA / EUAA spread – Typically, EUAAs are a less liquid market than EUAs as
the latter have great applicability. As a result, EUAAs tend to trade at a discount to
EUAs. Spread trades can be initiated if a trader believes that the differential is not fundamentally justified. For example, if the trader believed that the discount is too high
(i.e. EUAAs are trading significantly lower than EUAs), then the trader would buy the
EUAA and sell the EUA.
Option strategies – For those participants wishing to express views on volatility,
the range of ‘classic’ option strategies such as straddles and strangles could be utilised.
Volatility strategies were covered in Chapter 3.
11.8
WEATHER DERIVATIVES
‘I think it’s a perfect market. You can’t spook it, you can’t manipulate it. You
can’t make people think it’s going to be 110 degrees in London next week,” he
says. “And of course, weather is absolutely uncorrelated [to other asset classes]’.
—Financial Times (2007)
The common misconception relating to the difference between weather and climate was highlighted at the start of the chapter. Although weather derivatives have
been traded for many years, the market is relatively small and niche. So, what is weather
risk? Arguably from a corporate perspective it represents the potential adverse impact
of weather on cash flow and profits. The classic representation of this is given by the
impact of a cold summer spell on the sales of ice cream. However, other potential users
might include insurance companies, farmers, or ski resorts. Quantifying their exposure
to weather is anything but trivial.
The first weather derivative was attributed to a 1997 swap transaction between
Enron and Koch. Under the terms of the deal, Enron would pay Koch USD 10,000 for
every degree the temperature fell below a predetermined level; Koch would pay the
same for every degree above it.
It may be tempting to think of a weather derivative as a form of insurance. The
essence of insurance is to cover high-risk, low-probability events such as hurricanes
or floods. On the other hand, weather derivatives protect revenues against low-risk,
high-probability events such as mild winters that would reduce heating demand or wet
389
Climate and Weather
summers. With insurance the claimant must prove that they have suffered some form
of financial loss in order to be compensated. However, with derivatives an independent
third party will establish whether a payout is due under the terms of the contract.
11.8.1
Potential industries
Examples of potential users include:
▪ Agriculture
▪ Aviation
▪ Construction
▪ Energy
▪ Film production
▪ Retail
▪ Tourism
▪ Wholesale
11.8.2
General characteristics
Some common weather derivative characteristics include:
▪ An underlying weather index or variable.
▪ Temperature (average temperature, heating degree days, cooling degree days).
▪ Precipitation (rainfall, snowfall).
▪ Apparent temperature (wind chill, heat index).
▪ A period over which the index is measured (e.g. month, season, or year).
▪ A weather station to report the underlying variable (e.g. London Heathrow).
▪ A dollar value for each index point move such that either the index or the variable
can be ‘monetised’.
▪ A strike price for the index, which may be based on some historical value.
11.8.3
Exchange traded futures
The CME has traded weather futures for many years. The concept of Heating Degree
Days (HDD) and Cooling Degree Days (CDD) were introduced in the electricity chapter
(Section 8.4). These measures relate to the energy required to heat a space such as a
room or a building. The metrics are expressed relative to some benchmark value such
as 65∘ F. For actual temperatures below this benchmark, the space will require heating,
while above this level it will require cooling. It is tempting to think that HDD refers to
increases in temperature, but it is the opposite; the same is true of CDD.
In relation to exchange traded futures, the definitions for each term is:
▪ HDD index – This index is based on the sum of the average degrees that the
outside air temperature drops below the base temperature of 65∘ F (18∘ C for
non-US cities) in the specified city each day in the contract month. This could be
written as:
HDD = SIGMA MAX (0,65∘ F − Actual temperature)
390
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
▪ The HDD index will increase as the actual temperature falls relative to the stated
average of 65∘ F.
▪ CDD index – This index is based on the sum of the average degrees that the outside
air temperature rises above the base temperature of 65∘ F (18∘ C for non-US cities)
in the specified city each day in the contract month. This could be written as:
CDD = SIGMA MAX (Actual Temperature − 65∘ F, 0)
▪ The CDD index will increase as the actual temperature relative to the stated average
of 65∘ F increases.
▪ Reference is sometimes made to a Cumulative Average Temperature (CAT) index.
This index is based on the sum of the average temperatures recorded in the specified
city for each day in the contract month.
As an example of a futures contract consider the following contract specification:
Contract name:
Contract unit:
Price quotation:
Minimum price change:
Settlement:
Reference city:
US monthly weather Cooling Degree Day
USD 20.00 times the respective CME Degree Days
Index
Dollars per index point
One index point
Cash-settled
Dallas, Texas
By way of illustration, suppose that the futures index value for the month of August
is quoted at 691. To calculate the index, it is necessary to record the mean air temperature each day. If it was 100∘ F in Dallas on a particular day the index would be recorded as
35 (100∘ F − 65∘ F). If the temperature recording returned a value of 60∘ F then the index
would take a value of 0. Typically, these calculations are done daily and then summed
to determine a monthly or annual value. Note that the value used in this example was
based on a quote given in March of the same year so it suggests that traders will require
some form of model to predict the weather. The most known models are:
▪ The Global Forecast System (GFS) from the USA.
▪ The European Centre for Medium Range Weather forecasts (ECMWF) from
Europe.
The futures methodology then assigns an arbitrary fixed value of USD 20.00 (‘tick
size’) to each of those index points. The absolute value of this tick size is to an extent
irrelevant as it is the same for all the participants. So, the August contract analysed earlier will have an associated monetary value of USD 13,820 (691 × USD 20.00). Readers
familiar with equity derivative index futures will recognise this method of ‘monetising’
an index value.
391
Climate and Weather
11.8.4
Over-the-counter structures
Some typical characteristics seen in OTC trades are:
▪ Deals can be structured to cover a wider array of underlying exposures such as precipitation, snowfall, wind speed, and sunshine hours. However, transactions that
reference temperature are the most popular.
▪ May be able to accommodate any basis risk. That is, there could be a difference
between the weather at an official station and the affected site. Arguably this is
more relevant for precipitation than temperature.
▪ They will be traded for monthly, season, or annual periods.
▪ HDDs for winter months would typically be defined as October–April, while CDDs
for summer months would cover the May–September period. Autumn would be
defined as September–November, and Spring as March–June.
▪ The transaction will reference a given city (e.g. Dallas, Texas) or weather station (e.g. London Heathrow). Most weather observations are sourced from
government-run weather stations that follow standards set out by the World
Meteorological Organisation.
▪ Weather data is usually ‘cleaned’ as they sometimes contain errors or missing data.
There is no single market-wide technique used to clean data.
▪ Each transaction will have an associated tick value, which will determine the magnitude of the payout. In the Enron/Koch swap example cited earlier this was USD
10,000. The tick value is negotiated between counterparties but could be 1,000, 500,
or 200 in the respective currency.
▪ Transactions may be relatively short dated (e.g. current month or month ahead).
▪ Payments may be capped as, theoretically, there is no maximum or minimum value
for an observed temperature.
11.8.5
Swaps
A possible swap transaction could be:
Underlying commodity:
Trade date:
Effective date:
Maturity:
Tick value:
Fixed index value:
Floating payment:
Cooling Degree Days
1 March
1 August
One month (e.g. the month of August)
USD 500.00 per index point
691
Average of daily observed CDD index for Dallas, Texas
The settlement amount for this swap could only be calculated in the month
that follows the chosen period. To calculate the floating amount, the transaction
participants would need to collect a daily value of for the index, which will then be
averaged to calculate a settlement value. Suppose that the observed CDD value for the
392
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
month of August turns out to be 700. The difference between the observed and initial
fixed index value is 9 points. With a tick value of USD 500.00 per point this would
equate to a payment of USD 4,500 by the floating payer.
11.8.6
Options
From an option perspective both vanilla and exotic structures have been traded.
The payoff of a call option referencing weather would be:
Dollar value (tick value) of an index point × MAX (Index level − strike, 0)
The payoff of a put option referencing weather would be:
Dollar value (tick value) of an index point × MAX (Strike − Index level, 0)
The exotic structures have tended to be binary options where the option buyer
would receive a fixed payoff if the option is exercised, irrespective of how far above
the option is in the money.
11.8.7
Applications – cattle industry
Suppose a cattle farmer has analysed the impact of weather on his livestock and concludes that he has an exposure to below average temperatures in the autumn and above
average temperatures in the spring. For example, excessive heat may limit the growth
of animals in the spring, while younger animals may suffer if the temperature is very
cold. Weather could have an indirect impact on cattle costs as it may impact crop yields
which would influence the price of animal feed. One solution using options would be:
▪ The purchase of a call option on a CDD index for the spring period. As the temperature rises above the agreed strike, business losses will be offset by a payout on the
option.
▪ The purchase of a call option on a HDD index for the autumn period. For every
index point that the temperature falls below strike in the fall the farm will receive
a payout on the call option.
Readers may reasonably question why a call option rather than a put option has
been used for the HDD structure. Recall that any HDD index will increase in conditions of relatively cold weather. If the actual recorded temperatures in Fahrenheit over
a one-week period were 60, 55, 58, 61, 59, 56, 62 then the HDD index would be 44
(5 + 10 + 7 + 4 + 6 + 9 + 3). So, as it becomes colder the realised HDD index would
increase. Using these figures, the buyer of a one-week call option with a HDD index
strike of, say, 0 and an agreed dollar value per index point of USD 200.00, would receive
a payout of USD 8,800 (USD 200.00 × 44).
The strikes could be set either based on historical average temperatures or
around temperatures that the farmer considers optimal for feeding purposes. If the
farmer believed that in the autumn period (91 days) losses would be incurred when
temperatures were below 40∘ F he may choose a strike of 2,275 ((65−40) × 91) for the
period.
Climate and Weather
11.8.8
393
Applications – power utilities
A power utility in the US mid-west has determined that they have exposure to weather
risk in the summer. This is mainly driven by the demand for air conditioning units so
when it is cool there is less demand for electricity and therefore profits will fall. As a
result, they decide a hedge that references an average temperature would be appropriate. They determine that for every 1∘ F the temperature is below the current long-term
average, they will lose USD 50,000. There could be several different ways to structure
the transaction:
Buy a put option – This would reference the average temperature with the strike set
at an appropriate level that suits the client’s tolerance with respect to the upfront premium payment, the amount of downside risk and their ability to benefit from temperature increases. For example, an OTM put option, which reduces the premium would
have the strike level set below the long-term average temperature.
Zero premium collar – This would be structured as the purchase of an OTM put
along with the sale of an OTM call. The put option would provide the required downside protection and the strike of the sold call would be set at such a level that the
upfront payment would be eliminated. It would mean that the client would not be able
to benefit from an increase in temperatures above the strike of the call. This type of
structure is also referred to as a min-max transaction and was covered in greater detail in
Chapter 5.
Swap structure – This would have the effect of fixing the company’s exposure to
temperatures to a given level. There would be no downside risk if temperatures fell,
as they would receive under the swap but neither would there be any upside participation, as the swap would require the client to make payments under the terms of the
transaction.
Again, since there is theoretically no maximum or minimum value for temperature,
each derivative component may incorporate a limit on the payouts in either direction
(e.g. USD 300,000).
CHAPTER
12
Agriculture
12.1
AGRICULTURAL MARKETS
When discussing this segment of the commodity market it is often convenient to distinguish between two main sub-categories. Soft commodities include products such as
coffee, sugar, cocoa, and cotton. The production of these commodities is often concentrated in a small group of developing countries and as such can be prone to output
problems. Agricultural commodities are typically widely produced in developed countries. For example, wheat is widely produced and as a result farmers can usually respond
quickly to rising prices by expanding their acreage.
12.2
DEFINITIONS
As ever, it is always useful to define some relevant terms:
Supply – this usually comprises of three main components:
▪ Surplus stocks left over from the previous year. This could be referred to either as
‘inventory’ or sometimes ‘carry in’.
▪ Production from the current year.
▪ Imports.
Demand – typically comprises of two elements:
▪ Domestic use
▪ Exports
Carry over – this is defined as the remaining supply from the previous year plus
current year production plus imports.
Stocks to use ratio – this is defined as the current year ending stocks divided by
current year use. As a rule of thumb, in ‘normal’ markets the ratio should be 20–40%.
A lower ratio could indicate market ‘tightness’ and possibly increased price volatility.
394
395
Agriculture
12.3
AGRICULTURAL PRODUCTS
12.3.1
Physical supply chain – wheat
The physical supply chain (Figure 12.1) for grains such as wheat would typically start
with private farmers who would grow the crop. This source of supply could either be
domestic or could comprise of imports. The farmers would store this temporarily in
their own silos before selling it to a larger commercial trader referred to generically as
a grain elevator (although strictly speaking a grain elevator is a term used to describe a
building used to store grain before onward shipment). These larger commercial traders
(e.g. Cargill) could accumulate and combine the production of several smaller farmers.
They may also perform activities such as inspecting, cleaning, and blending. These commodity trading houses in the agricultural markets are dominated by four main entities,
referred to as the ABCDs.
▪ Archer Daniels Midland Company
▪ Bunge Ltd.
▪ Cargill Inc.
▪ Louis Dreyfus Co.
The next stage is milling, whereby the wheat is ground and sifted to make flour
for human consumption and mill feeds, which are sold as animal feed. The final stage
is the baking process, which would produce breads and cakes. Flour can also be used
in the production of pasta-style products or for home baking. This part of the market
would include well-known household names such as General Mills, Kellogg, and Kraft.
The final product would then be sold to the end consumer either through a shop or
restaurant.
Bakery
Retail outlets
Farmers
Consumer
Merchant /
elevator
Flour mills
Other
manufacturers
Imports
Animal feed
FIGURE 12.1 Simplified physical supply chain for wheat.
Restaurant
396
12.3.2
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Wheat
There are several varieties of wheat, but its main use is in the production of flour,
which is then used in foods such as bread and pasta. Although there are many different
‘species’ of wheat, the US produces six different types:
▪ Hard Red Winter – used for milling baking flour, Asian noodles, hard rolls, flat
breads, tortillas, and general-purpose flour.
▪ Hard Red Spring – sometimes referred to as the ‘aristocrat’ of wheat used for rolls,
croissants, bagels, and pizza crusts.
▪ Soft Red Winter – used in cookies, crackers, pretzels, pastries, and some flat breads.
▪ Durum – pasta, couscous, and Mediterranean breads.
▪ Hard White – similar uses to hard red winter.
▪ Soft White – cakes and pastries.
The spring/winter classification relates to when they are planted with most of the
production being based in the Midwest and eastern states.
Figures used in the first edition of this text indicated that global production for
2006–2007 was approximately 585 million tonnes, while consumption for the same
period has been estimated at 613 million tonnes. By 2020 the figures were 764 million
and 754 million, respectively. Both figures indicate that production and consumption
are rarely equal indicating that the level of inventories is a key price driver.
In terms of production by region (Figure 12.2), the EU accounts for 26% of the global
total. The second largest producer is China at 23%, followed by India (17%), Russia (15%)
and United States (8%).
In term of consumption by country (Figure 12.3), the single biggest consumer of
wheat is the EU, which consumes 21% of all wheat. China accounts for 20% of the global
total, India accounts for 16%, while the figure for the USA is 5%.
12.3.3
Corn
Traditionally corn (sometimes referred to as maize) has been used as a feed for livestock. However, it has applications for human consumption such as sweeteners for soft
drinks. However, the demand and supply dynamics of this product can be impacted by
the price of oil, as farmers may be tempted to direct their output towards the production
of ethanol.
Global production of corn in 2006–2007 was estimated at 689 million tonnes, while
the figure for consumption was 724.1 million tonnes. Like wheat, both production and
consumption has increased significantly so that by 2020 the figures were 1,111 million
tonnes and 1,135 million tonnes, respectively.
The largest producer of corn is the USA (Figure 12.4) accounting for about 34% of
the global total. China is the second largest producer (24% of the total), with the next
largest being Brazil (7%) and the EU (6%).
The United States consumes about 28% of global production (Figure 12.5), while
the respective value for China is 25%, the EU is 7%, and Brazil is 6%.
397
Agriculture
USA
52,258
EU
154,000
Russia
73,500
India
102,190
China
133,590
U.S. Department of Agriculture
FIGURE 12.2 Wheat production (thousand metric tons).
Source: USDA, www.ers.usda.gov
USA
31,706
Russia
39,500
China
128,000
India
98,000
European Union
127,000
FIGURE 12.3 Wheat production (thousand metric tons).
Source: USDA, www.ers.usda.gov
398
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Argentina
50,000
European Union
65,000
Brazil
101,000
USA
347,782
China
260,770
FIGURE 12.4 Corn production (thousand metric tons).
Source: USDA
Mexico
44,500
Brazil
66,500
USA
313,578
European Union
82,500
China
279,000
FIGURE 12.5 Corn consumption (thousand metric tons).
Source: USDA
399
Agriculture
12.3.4
Palm oil
Palm oil is one of the major edible oils and fats that are traded on a global basis. Other
competing products include soya bean oil and rapeseed oil. Palm oil can be used for a
variety of purposes reflecting the different applications of agricultural products:
▪ Food products – cooking oils, shortening or dough fats, and margarine.
▪ Chemical feedstock – used in cosmetics, toiletries, industrial cleaning agents, and
candles.
▪ Fuel – can be used in the production of biodiesels.
Most palm oil (70%) is used within the food industry with the remainder (30%) used
for industrial purposes (Figure 12.6).
Production is dominated by Indonesia and Malaysia who make up nearly 85% of all
production (Figure 12.7). This is because oil palms can only be grown in countries with
a tropical climate that is 10∘ N or 10∘ S of the Equator.
The major importers are India, the European Union, and China (Figure 12.8).
There are three main types of oil palm: Dura, Pisifera and Tenera, the latter of which
is the most used in commercial production. The production cycle is relatively short
and so the production response to higher prices is relatively quick. There is a nursery stage of 12–18 months before planting with the first harvest and can be harvested
around 30 months later. The oil palm will have a lifespan of about 25–30 years with
peak production in years 8–15. Since the plant is grown in a tropical climate it can be
harvested all year round. Barclays (2011a) estimates that as a rule of thumb, one hectare
can produce about four tons of palm oil although this will vary depending on the fertility
and toxicity of the land and the incident of diseases.
80
70
60
50
40
30
20
10
0
2007
2008
2009
2010
2011
2012
2013
Industrial use
2014
FIGURE 12.6 Usage of palm oil (millions metric tons).
Source: USDA
2015
Domestic use
2016
2017
2018
2019
2020
400
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Thailand
3,000
Colombia
1,680 Nigeria
1,015
Malaysia
19,800
Indonesia
42,500
FIGURE 12.7 Production of palm oil (thousand metric tons).
Source: USDA
Philippines Egypt
1,150
USA 1,220
1,550
Bangladesh
1,700
India
9,750
Pakistan
3,350
China
7,200
FIGURE 12.8 Importers of palm oil (thousand metric tons).
Source: USDA
European Union
7,300
401
Agriculture
12.3.5
Soybeans
Soybeans are believed to have originated in China before appearing in Europe about
300 years ago. They are generally planted in the spring (April and May), but can be
planted as late as July. If the crop is planted later in the year there is a possibility
that it could be damaged by frost, which would reduce supply and potentially lead to
higher prices.
Soybeans are classified as an oilseed, a category that also includes peanuts, sunflower seeds, and rapeseed.
The first step in the production lifecycle is the removal of the outer skin of the
bean, which is referred to as the soy hull. This can be used as feed for the dairy industry. The bean could then be consumed as food or crushed (with or without the hull) to
produce soybean meal and soybean oil. Christian (2006) argues that this crushed material is sometimes used as a proxy for demand although not fully accurate as it does not
reflect the fact that some beans are consumed without crushing and other beans are
used as seeds for the following year. Soybean meal is used as a feed for livestock while
soybean oil can be used for human consumption (soy milk, oils, and dressings) as well
as applications as diverse as soap and explosives.
Figures 12.9, 12.10, and 12.11 show the production of soybeans, soybean meal, and
soybean oil; the so-called ‘soybean complex’.
China
18,100
Argentina
54,000
Brazil
126,000
United States
96,841
FIGURE 12.9 Production of soybeans (thousand metric tons).
Source: USDA
402
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Argentina
33,525
China
68,112
Brazil
33,950
United States
44,881
FIGURE 12.10 Production of soymeal (thousand metric tons).
Source: USDA
Argentina
8,500
China
15,411
Brazil
8,400
United States
11,018
FIGURE 12.11 Production of soy oil (thousand metric tons).
Source: USDA
403
Agriculture
12.4
SOFT COMMODITIES
12.4.1
Sugar
According to the International Sugar Organization (2020):
▪ About 110 countries produce sugar.
▪ Sugar can be produced from either cane or beet, but production is dominated
by cane, which accounts for about 80% of output. Beet tends to be grown in the
Northern Hemisphere while cane is produced in the Southern Hemisphere.
▪ The main producing countries are India, Brazil, Thailand, China, USA, Mexico,
Russia, Pakistan, France, and Australia. These countries account for about 70% of
output.
▪ Sugar can be used for food, animal feed, and fuel (sugar-based ethanol).
▪ By 2018 the annual consumption of sugar was just over 172 million tonnes.
Although overall consumption continues to increase, in recent times the rate of
growth has decreased.
▪ The main consuming markets are India, EU, China, Brazil, USA, Indonesia, Russia,
Pakistan, Mexico, and Egypt.
▪ Approximately 64 million tonnes of sugar are traded internationally. Brazil,
Thailand, the EU, Australia, and India accounted for 70% of exports with the
market being dominated by Brazil. The major importers are Indonesia, China, and
the USA.
12.4.2
Coffee
The International Coffee Organisation (ICO) defines coffee in a somewhat formal manner as ‘(a) general term for the fruits (cherries) and seeds (beans) of plants of the genus
Coffea, as well as products from these fruits and seeds in different stages of processing
and use’.
Coffee bean structure
When a coffee plant matures, the fruit, which contains the bean, will have either a red
or yellow external colour. Each plant will have several outer layers, which cover the
coffee bean. The next layer is the pulp, which consists mainly of water and sugar and
is removed and can be used as compost. The mucilage is a sweet coating, which can
be used to make jelly. The parchment is the last outer layer before the bean, which can be
transformed into either mulch or paper. The coffee bean (a.k.a. the seed) is protected by
a silver skin, which has applications as a cigarette filter.
Where is coffee grown?
Coffee is indigenous to Africa with arabica believed to have originated in Ethiopia
and robusta from the Atlantic coast. The bulk of coffee is produced in Latin America
with Brazil the dominant grower. The other major producers are Vietnam, Colombia,
404
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Indonesia, and Honduras. It is also common to refer to either the ‘bean belt’ or ‘coffee
belt’, which is a region that sits between the Tropics of Cancer and Capricorn where the
environmental conditions for growing coffee are ideal and encompasses the majority
of production.
Types of coffee
There are two main types of coffee bean that are traded internationally: arabica and
robusta. Two other types exist (excelsa and liberica) but these are only produced in relatively small quantities. Internationally traded coffee is often referred to as beans, which
is a commercial term used to describe the dried seed of the coffee plant. Arabica is often
perceived as having a smoother flavour while robusta is relatively bitter and often used
for instant coffee.
The ICO reports four categories of coffee from different countries each group having
distinctive characteristics:
▪ Brazilian natural arabicas – coffees from Brazil, Ethiopia, Paraguay.
▪ Colombian mild arabicas – Colombia, Kenya, Tanzania.
▪ Other mild arabicas – made up of 22 countries such as Costa Rica, Peru, and
Rwanda.
▪ Robustas – made up of 23 countries such as Ghana, Indonesia, and Vietnam.
Reference is sometimes made to ‘specialty coffee’. Arguments abound as to how this
is actually defined but the Specialty Coffee Association (SCA) describes it as ‘coffee that
met all the tests of survival encountered in the long journey from the coffee tree to the
coffee cup’. So, this perspective does not look exclusively at the quality of the bean but
the entire supply chain process. Readers interested in the technical standards applied
by the SCA are referred to their website.
Overview of production process
The plant will first produce a flower followed by the fruit, which contains the final bean.
Each plant will take about three to four years from first planting to produce coffee and
with careful management will last about 25 years. The fruit is then harvested, washed
to remove any dirt or stones, and then dried for one to two weeks to remove as much
moisture as possible. The beans are then packaged into sacks and sent to a warehouse
where they will typically spend between six months to a year maturing to improve their
flavour. Some producers will mature them for up to three years as they will keep quite
well if they are completely dry.
Roasting is done at temperatures of 190–250∘ F. The longer the roast time, the
darker and more flavorsome the coffee becomes, and vice versa. Some roasting
examples include:
▪ 8 minutes – Light roast
▪ 12 minutes – City roast (style of roast preferred in the US)
Agriculture
405
▪ 20 minutes – Italian roast
▪ 24 minutes – French roast
▪ 26 minutes – Extreme dark
Once roasted the beans are ground and are then ready to be brewed into the final
drink. As a rule of thumb, it takes about 10 lbs. of beans to produce 2 lbs. of coffee.
Amount of caffeine
The following types of popular drink contain different amounts of caffeine:
▪ Decaf – 3mg
▪ Hot chocolate – 19mg
▪ Green tea – 20mg
▪ Shot of espresso – 27mg
▪ Can of cola – 40g
▪ Black tea – 45g
▪ Energy drink – 80mg
▪ Instant coffee – 82mg
▪ Brewed coffee – 95mg
Different forms of coffee
When discussing coffee, reference is often made to coffee in different forms at different
processing stages:
▪ Green coffee captures all coffee in the naked bean form before it is roasted.
▪ Dried coffee cherry means the dried fruit of the coffee tree.
▪ Parchment coffee means the green coffee bean contained in the parchment skin.
▪ Roasted coffee means green coffee roasted to any degree and can include ground
coffee.
▪ Decaffeinated coffee means green, roasted, or soluble coffee from which the
caffeine has been extracted.
▪ Liquid coffee means the water-soluble solids derived from roasted coffee and put
into liquid form.
▪ Soluble coffee means the dried water-soluble solids derived from roasted coffee.
Production and consumption statistics
The concept of a ‘marketing year’ is often referred to with respect to agricultural commodities. This is typically a period of one year where data such as production and
consumption trends are reported and analysed. This period may or may not coincide
with the harvest period but may be determined by wider market activity. Since coffee
is produced in different regions each will have their own harvest date and so the ICO
adjust all the data such that it is all reported on a common basis matching the marketing
406
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
year. For example, the International Coffee Organisation (ICO) has a marketing year
that commences on 1 October that attempts to capture most harvest dates.
Production is often described in terms of units of 60-kilogram bags. For 2019 the
ICO estimates production at approximately 168 million 60-kg bags, with consumption
being a little higher. Production of the two main coffee types was 96 million 60-kg bags
with Arabica accounting for an estimated 72 million 60-kg bags. From a geographical
perspective, the relative production and consumption values are shown in Figures 12.12
and 12.13.
Market structure
Figure 12.14 aims to capture in a simplified manner the different stages of the coffee
supply chain.
In many countries (e.g. Costa Rica) coffee is largely produced in small-hold operations although larger production facilities exist. A local co-operative could then act as a
single point of contact to collect the different sources to create an overall larger volume
that could be sold commercially. The beans could then be sold either to a trading house
acting as an intermediary or directly to a roaster. These roasters will then transform
the beans into different flavours of coffee by roasting and/or blending various beans.
Africa
11%
South America
46%
Asia & Oceania
30%
Mexico and Central America
13%
FIGURE 12.12 Global coffee production by region.
Source: International Coffee Organisation
407
Agriculture
Africa
8%
North America
23%
Asia & Oceania
26%
Mexico and Central America
4%
Europe
39%
FIGURE 12.13 Global coffee consumption by region.
Source: International Coffee Organisation
Smaller
producer
Co operatives
Larger
producer
Initial
processing
Initial
processing
Trading
house
Example
• Louis Dreyfus
Roaster
Retail
outlet
Example
Example
• Specialist
• Shop
• Retail coffee
• Coffee shop
shops (e.g. Starbucks) • Supermarket
• Commercial producers • Institutional
(e.g. Kraft)
End
product
Example
• Beans
• Ground
• Soluble
• Decaf
• Ready to drink cans
FIGURE 12.14 Simplified coffee supply chain.
Roasters range in size from small specialist operations to the very large commercial
producers. By way of illustration the following quote is taken from the 2019 Starbucks
annual report:
‘To ensure compliance with our rigorous coffee standards, we control substantially all coffee purchasing, roasting, and packaging and the global distribution
of coffee used in our operations. We purchase green coffee beans from multiple coffee-producing regions around the world and custom roast them to our
exacting standards for our many blends and single origin coffees’.
408
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The final product could then be sold in a variety of different retail outlets such as
branded coffee shops or supermarkets. The end retail product could consist of items
that vary from the simple whole bean to ready-to-drink coffee in a can.
Prices
Like many products the coffee market tends to rely on a futures price for the pricing of
commercial contracts. In its annual report Starbucks says:
‘Although most coffee trades in the commodity market, high-altitude arabica
coffee of the quality sought by Starbucks tends to trade on a negotiated basis
at a premium above the “C” coffee commodity price. Both the premium and
the commodity price depend upon the supply and demand at the time of purchase. Supply and price can be affected by multiple factors in the producing
countries, including weather, natural disasters, crop disease, general increase
in farm inputs and costs of production, inventory levels and political and economic conditions. Price is also impacted by trading activities in the arabica
coffee futures market, including hedge funds and commodity index funds’.
The ‘C’ coffee price refers to the futures price for arabica coffee traded on exchanges
such as the ICE. Like all futures exchanges it can be used as a source of supply, a risk
management and trading venue, as well as a pricing basis for commercial transactions.
The C contract references arabica beans of a specified grade from one of 20 countries.
Since no two beans will be perfectly identical any beans that will be physically delivered
are tested for grade and flavour. Grading can be based on criteria such as:
▪ Altitude or region
▪ Preparation
▪ Bean size, shape, colour
▪ Number of defects
▪ Flavour
The exchange uses certain coffees to establish a baseline (e.g. Mexico and Costa
Rica). Certain coffees are judged to be of a higher quality and are priced at a premium
(e.g. Colombia), while those judged inferior are priced at a discount (e.g. Brazil). Note
that the premium mentioned in the Starbucks quote relates to bilaterally negotiated
contracts and is therefore variable. The premium used in the futures market is a fixed
value established by the exchange, which is applied to the final invoice if physical
delivery were to take place.
The ICO also publishes a series of prices, which are designed to track the spot prices
of the four main coffee categories established by the organization. They also produce a
Daily Composite Indicator price that combines all these prices to give a single measure
that represents the current international coffee price.
12.4.3
Cocoa
According to the International Cocoa Organisation, cocoa is produced in countries in
a belt between 10∘ N and 10∘ S of the Equator, where the climate is optimal for growing
409
Agriculture
cocoa trees. The largest producing countries are the Ivory Coast, Ghana, and Indonesia.
In terms of consumption, the EU region followed by the USA account for most of the
demand. Within the market, consumption is not measured directly but is inferred from
the amount that enters the grinding stage of the process.
The plant can be harvested all year round but generally needs opposite conditions to
coffee, i.e. lower altitudes and higher temperatures. From its initial planting it will take
about 1.5 years for a plant to flower then another six months before the fruit (known
as pods which contain the cocoa beans), are produced. The outer husk of the plant can
be used as compost. Inside each pod there are several beans contained within a milky
fluid. This white fluid and the beans are then left to ferment for up to a week before
separation. The beans are then dried for one to two weeks to remove any moisture. The
beans are then roasted for about five minutes at 50∘ C to make them ‘crunchy’ such that
the outer shell can be easily removed. The inner ‘nibs’ are then ground to create cocoa
liquor, which has the consistency of paste. This paste is then pressed, which separates
into cocoa butter and ‘presscake’. The cocoa butter is used in the production of chocolate
and has applications in cosmetics (e.g. lotions) while the presscake is pulverised to form
cocoa powder.
There are three main types of chocolate:
▪ Dark – cocoa, sugar, and fat,
▪ Milk – cocoa, sugar, fat, and butter,
▪ White – cocoa, white sugar, fat, powdered milk.
The world cocoa market distinguishes between two broad categories of cocoa beans:
‘fine’ or ‘flavour’ beans and ‘bulk’ or ‘ordinary’ beans. In terms of market structure, at a
very high level it will resemble the illustration in Figure 12.4. The largest confectionery
company is Mars Wrigley, and they may wish to buy the beans directly from source
before processing and producing the final product. There are also smaller entities such
as Hotel Chocolat in the UK that specialises in the production and retail sale of ‘luxury’
chocolate produced from their own plantation in Saint Lucia.
12.5
ETHANOL
‘If you look at the ethanol landscape worldwide, the US and Brazil are like
Kuwait and Saudi Arabia’.
—Fabrizio Vichichi – Financial Times, 26 January 2006
12.5.1
What is ethanol?
Ethanol has the simplest chemical structure of the hydrocarbons within the family of
alcohols. In chemistry, alcohol is used to describe a hydrocarbon with an -OH group
attached to a carbon atom. Ethanol is just one of many alcohols, though the words are
sometimes incorrectly interchanged in the media.
When ethanol is produced for fuel, the goal is to maximise its strength and minimise
any impurities. As a result, the primary production inputs such as corn and sugar cane
410
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
are chosen for their ease of cultivation and their high carbohydrate levels. The energy to
input ratio for ethanol from sugarcane is 8.3 versus 1.9 for sugar beets and 1.8 for corn.
Initially the corn or other crop is mashed and then heated to destroy any bacteria
or other microorganism that may interfere with fermentation. The mash is then cooled
and yeast is added. The yeast feeds on the mash releasing carbon dioxide and ethanol.
Since the carbon dioxide is a gas, it bubbles out of the fermenting mash leaving the
ethanol behind.
At this stage, the ethanol contains many impurities, which, along with any
water, are removed by distillation. Any remaining impurities are removed by using a
molecular sieve.
Proponents of ethanol produced from sugar cane argue that when used as a fuel it
is environmentally friendly; the carbon dioxide produced from burning it is neutralised
by the carbon dioxide absorbed by the plants from which it is grown. However, counter
arguments point out that the increased demand for sugar cane in Brazil may encourage production to expand into environmentally sensitive areas such as the Amazon.
Critics also argue that some of the technology used to produce ethanol requires a substantial power input, which may offset any perceived environmental benefits. Other
issues relate to the use of fertilisers for the rapid and efficient growth of the sugar and
corn. Fertilisers are synthesized industrially from ammonia, which is made from nitrogen and hydrogen. The latter is obtained from hydrocarbon reserves with the inevitable
production of carbon dioxide, and so any ‘green’ benefits derived from ethanol may
be limited.
There are alternative biofuels that include:
▪ Cellulosic ethanol – uses a variety of other materials such as waste plant material
to produce the ethanol, without as many side effects.
▪ Biobutanol – produced from sugars like ethanol and has a higher energy content.
Can be used in higher concentrations in existing car engines.
▪ Biodiesel from energy crops – this is biodiesel produced from oil from plants that
can be grown on land not suitable for food production.
▪ Biomass to liquids – Biodiesel produced by taking gas from decomposing vegetation
and turning it into a liquid.
At the time of the first edition of this text, world production in 2004 for ethanol was
estimated at 10.7 billion gallons (approximately 49 billion litres) and 12.1 billion gallons
(55 billion litres) in 2006. (Financial Times, 9 May 2006). By the end of 2018, production
was approximately 24 billion gallons (108 billion litres).
12.5.2
History of ethanol
According to the International Sugar Organisation (2020), ethanol for fuel can be used
in two ways:
▪ It can be blended with gasoline to reduce petroleum use, boost octane ratings, and
cut exhaust emissions.
▪ ‘Pure’ ethanol, defined as a fuel made up of 85–100% ethanol, which can be used in
specially designed engines such as flexifuel vehicles.
411
Agriculture
Arguably the first significant use of ethanol as a biofuel was in Brazil in the 1970s.
At the time, Brazil imported 90% of its crude oil so the government acted to reduce
the impact of rising prices. Through a series of subsidies to sugar mills, ethanol was
promoted as an alternative fuel source. By the 1980s, new car production was focused
almost exclusively on cars that ran purely on ethanol. However, this first boom was
short lived because of a combination in factors:
▪ The country struggled to produce sufficient sugar cane to meet demand.
▪ The government lifted controls on the pump price of the fuel, leading to an increase
in price.
▪ An increase in the global price of sugar, which encouraged producers to switch back
to using the raw material to make refined sugar.
▪ Mediocre performance of the vehicles.
▪ Subsequent price collapse of crude oil, making gasoline relatively attractive.
However, in 2003 the country introduced flexifuel cars, which operate either
on ethanol or petrol or a mixture of the two fuels. By 2006 estimates suggested that
50–75% of all domestic new car sales were of the flexifuel variety and that about half
of the country’s sugar cane crop was being used for domestic ethanol production.
As a result of their knowledge and experience gained in the 1970s and the relatively
low-cost nature of their ethanol operations, Brazil became known as the ‘Saudi Arabia
of biofuel’. However, based on current statistics, the USA produces about 56% of all
ethanol, while Brazil produces just 28%.
In the USA, about 40% of the annual corn crop is processed into fuel ethanol (Financial Times, 2019a). Traditionally, the government has offered a range of incentives (e.g.
government mandates, tax relief, import tariffs) along the supply chain to encourage
production and consumption. The demand for ethanol in the USA was boosted when
the government banned the gasoline additive MBTE (methyl tert-butyl ether) for environmental reasons. As a substitute, ethanol is now added to gasoline in concentrations
of between 10% (so-called E10 fuel) to 15% (E15). Beyond this level most normal internal
combustion engines experience operating difficulties unless they have been specially
converted.
In Europe, ethanol is produced from sugar beet (12.1 million tonnes), maize
(4.1 million tonnes), wheat (3.9 million tonnes), barley (0.4 million tonnes) and rye
(0.4 million tonnes).
12.5.3
Supply chain: corn to ethanol
Corn kernels are first delivered to a processing plant, where they are stored before
entering the production cycle. The kernels are then ground into a fine powder called
meal. This meal is mixed with water and enzymes that converts starch to sugar.
This liquid mixture is now called ‘mash’ and is heated to reduce levels of bacteria.
The mash is then cooled, and yeast is added to ferment the sugars into alcohol and
carbon dioxide. The alcohol is then separated by means of distillation. A molecular
sieve removes any remaining water, and the resulting ethanol can then be stored
or shipped.
412
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
One of the main stumbling blocks to the global usage of ethanol as a source of fuel
for transport is the cost of developing a retail distribution network. Additionally, questions have been raised about how feasible it would be to produce ethanol in sufficient
quantities to replace the current reliance on gasoline as the amount of land required
to produce sufficient quantities would be enormous. This could influence the amount
of land dedicated to grow corn for feed, which would impact other parts of the food
supply chain. There is also the related issue of moving the ethanol to the end point of
consumption. In most countries the main method is to move it by trucks or train. It is
not feasible to move it using the existing hydrocarbon pipelines, as water tends to build
up in them, which would attach it to the ethanol, watering it down and rendering it
useless as a fuel.
12.6
PRICE DRIVERS
There are three main themes commonly discussed within the context of agriculture.
They are the three ‘Fs’ of food, feed, and fuel.
By way of illustration, consider the following short quiz. What commodity links the
following situations?
▪ A sugar cane farmer in Brazil.
▪ A corn farmer in the USA.
▪ Increase in demand for meat in China.
▪ Increase in the price of soybeans in Iowa.
When the author asks this question in a classroom, typical answers include water,
land, fertiliser, transport, and climate. Some participants manage to identify the correct
answer, which is crude oil. The linkage? Rising crude oil prices would make it more
attractive for Brazilian producers to make more ethanol since it is a form of fuel substitute. Ethanol can also be produced from corn and so this might encourage US farmers
to divert more of their crop to this end use. However, corn can also be used as a feed for
cattle and as incomes in China increase there will likely be an increase in the demand for
meat. With a large proportion of corn being diverted into ethanol production, there will
be less corn for animal feed and so its price will increase. If US farmers respond by planting more corn, then there is a smaller area of land available for producing soybeans.
Since the supply of this product now falls the price would rise.
12.6.1
Physical market factors
Weather and climate change
A strong harvest will increase the amount of the product coming to the market. In this
respect the key factor is weather. A very hot summer, or indeed a very cold one, may
have an adverse effect on the size of the harvest. The impact of climate change will
influence long-term trends and is addressed in Chapter 11.
Agriculture
413
One significant aspect of weather is the El Niño Southern Oscillation (ENSO),
which is a climate pattern that is used for seasonal forecasting several months into the
future. This is a natural phenomenon that occurs from the interaction of the ocean
and atmosphere. ENSO comprises of two extremes El Niño (‘the boy’) and El Niña
(‘the girl’), which are, respectively, the warm and cold-water phases of the climate
pattern. They were so named by South American fishermen who, every few years,
would observe the conditions, which usually peaked around Christmas time.
These two phases are cyclical environmental patterns that occur across the Equatorial Pacific Ocean. In neutral years, the trade winds will blow from east to west and as
a result warmer water is pushed towards the eastern shores of Indonesia and Australia.
These warmer sea temperatures lead to warmer air rising, which can lead to more cloud
and unsettled weather. Off the coast of South America, as the warmer water is pushed
away, colder water will start to rise to the surface. This split between warmer weather in
the west and colder weather in the east leads to atmospheric circulation, which pushes
cooler dryer air towards the east.
El Niño is associated with a weakening of the trade winds which reduces the temperature differential as the warmer waters are not pushed as far west as a neutral year.
As a result, the eastern waters off South America become relatively warmer. This change
in weather patterns will have an impact on temperatures and rainfall.
La Niña is the opposite of El Niño and describes a situation where the trade winds
would be stronger than a neutral year. This pushes the warmer waters further westwards
and as a result the waters off South America become colder and extend further into the
Pacific.
El Niños typically occur every three to five years, and although there are some
common characteristics shared by each event, no two occurrences will be identical. According to the National Oceanic and Atmospheric Administration’s website
(https://www.noaa.gov/education/resource-collections/weather-atmosphere/el-nino),
the most common effects of El Niño are:
▪ Deficient rainfall over Indonesia and northern South America.
▪ Excess rainfall in the southeastern South America, eastern equatorial Africa, and
the southern US.
La Niña is the opposite pattern. For example, in the USA there will be drier weather
in the South, but the North West will be colder and wetter than average. Whereas El
Niño is associated with a reduction in hurricanes that form in the Atlantic, La Niña
events tend to be related to an increase.
The somewhat obvious conclusion is that ENSO’s influence on rainfall and
temperature can have a significant impact on the production and availability of food.
Physical constraints and restraints
Supply for a particular product may reach a peak if the country runs out of land
that can be used for production, or there is no labour or water available. There may
also be environmental and political pressures that could restrain the expansion of a
particular product.
414
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Transportation issues
A significant proportion of agricultural products that are exported will be done so by
seaborne means. Several ‘choke’ points exist where political events may result in a disruption to supplies. These include:
▪ Panama Canal
▪ Strait of Malacca
▪ Strait of Hormuz
▪ Bosphorus Strait
Cost of production inputs
Although supply chains will vary amongst different sectors an increase in the
cost of inputs such as diesel, fertilizer, and seeds will ultimately feed through to
higher prices.
Innovation and investment
One of the more popular examples of innovation has been in the coffee market with
the introduction of single serve capsule machines. Although in the long term it is
expected that this may attract new consumers into the market and increase demand,
the initial impact was the opposite. A common refrain was that a lot of coffee was
often ‘consumed by the sink’, but the single serve capsules removed this waste,
reducing demand.
If a particular product is faced with challenges to increase supply, it may require
some form of innovation. This could encourage better plant breeding of development
of higher yielding seed material. Genetically modified crops are now very common and
result in greater crop yields and more effective resilience to disease. However, not all
areas of the world will accept such grains and this type of investment often requires a
significant capital outlay.
Another form of investment is that some large end consumers will often work with
small hold growers of certain crops, particularly in emerging markets. They may provide
expertise and perhaps funding to secure the quality and quantity of the raw materials
that they need.
Crop switching
Following the sharp rise in energy prices during the early part of the century, the
interest in ethanol as a fuel substitute had a subsequent effect on the demand for
its primary production inputs, which include sugar, corn and soybeans. In theory as
crude oil prices increase, the demand for ethanol from sources such as corn becomes
more economical to produce. As a result, farmers will plant more of the crop that is
in demand and less of the lower demand crop. This is sometimes dubbed ‘the battle
for acreage’.
415
Agriculture
A related biofuel is biodiesel, which can be produced from renewable resources
such as soybeans or palm oil. Biodiesel has several advantages that make it attractive:
▪ It can be used in existing diesel engines.
▪ It can be blended with normal diesel.
▪ It is environmentally very friendly, emitting a lower amount of carbon than
conventional fuels.
Elasticity of supply
In the energy and metals chapters it was noted that the supply side response to
an increase in demand was relatively slow due to the time needed to construct the
required infrastructure. However, with respect to the agricultural sector, the response
time is determined by the relatively short period it takes to plant and harvest a
particular product.
Product substitution
One practice used in the coffee industry is for roasters who deal in less high-end coffee
to add more robusta to blends at times when arabica prices are high.
Current levels of inventory
The level of stockpiles (‘carry over’) is an indication of the balance between demand
and supplies and as such will have an impact on price. Another popular term that
is used in the market is ‘stock to use’ ratio. This indicates the amount of inventory carried over into a new reporting year as a percentage of the total use of the
commodity.
12.6.2
Societal factors
Health
The demand for certain soft commodities such as sugar may also be affected by a growing awareness of health-related issues. For example, a change in the way that food is
packaged may result in a change in dietary habits if the amount of sugar is made more
visible. Additionally, the development of sugar substitutes such as stevia may also result
in a fall in demand. Stevia is a plant and is 300–400 times sweeter than sugar, but has
no calories and does not raise blood sugar levels.
Population growth
In very simple terms a greater population means more mouths to feed and so in theory
greater demand. At the time of the publication of the first edition of this book (2007),
the world population was about 6.7 billion. At the time of writing some 13 years later,
it is now nearly 7.8 billion (Figure 12.15).
416
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
9,00,00,00,000
8,00,00,00,000
7,00,00,00,000
6,00,00,00,000
5,00,00,00,000
4,00,00,00,000
3,00,00,00,000
2,00,00,00,000
1,00,00,00,000
19
5
19 0
5
19 2
5
19 4
5
19 6
5
19 8
6
19 0
6
19 2
6
19 4
6
19 6
6
19 8
7
19 0
7
19 2
7
19 4
7
19 6
7
19 8
8
19 0
8
19 2
8
19 4
8
19 6
88
19
9
19 0
9
19 2
9
19 4
9
19 6
98
20
0
20 0
0
20 2
0
20 4
0
20 6
0
20 8
1
20 0
1
20 2
1
20 4
1
20 6
18
-
FIGURE 12.15 World population 1950–2018.
Source: Food and Agricultural Organization of the United Nations (FAO).
Holiday periods and religious events
Agricultural product prices may exhibit some form of seasonality depending on the time
of the year.
▪ Christmas and US Thanksgiving – Prices of turkeys will increase, generally peaking
in November. Fish related products such as smoked salmon and lobster will tend
to be high in the October to December period.
▪ Religious holidays – At Easter time approximately 80 million chocolate eggs are
sold in the United Kingdom (population approximately 67 million). As a result,
cocoa butter tends to rise in the first half of the year and then fall back. Ramadan
is associated with higher consumption of meats such as lamb.
12.6.3
Governmental intervention
Quality of data
It would be virtually impossible to be able to count every single agricultural product cultivated in a particular country on a real-time basis. Consequently, reporting agencies use
a variety of different sampling techniques to obtain the data. For example, one key report
published by the US Department of Agriculture is the ‘Prospective Plantings’ report,
which is based on a sample of about 80,000 farm operators from an estimated population of two million farms. Although the process is often very rigorous it is unlikely to
be perfect and revisions to data in some (but not all) reports can occur.
The Financial Times (2010f) reported that some trading houses would often
employ staff to stand at the gates of every cocoa warehouse in Abidjan, Ivory Coast,
to count the movement of their of competitor’s trucks to get a better picture of the
country’s output.
Anecdotally, the authors recall doing some work at an agricultural hedge fund
where the fund managers would often charter helicopters to see how well various crops
were growing.
417
Agriculture
Protectionism
Inevitably when considering the nature of agricultural commodities, the issue of tariffs and subsidies will arise. Subsidies will influence where and how much is produced,
while import tariffs may be designed to protect a domestic industry for social or political reasons. For example, acreage switching due to differing grain subsidies is a factor
monitored by some traders. More extreme measures may include limits on production,
a ban on food exports (to keep food in domestic markets), or a ban on imports (to protect local producers from cheaper imports). Sometimes, export tariffs may encourage
consumers to switch new sources of supply in a different geographical region.
Subsidies to farming usually take the form of either an income supplement or a minimum price guarantee. This has led to international trade disputes such as those seen
between the USA and Europe. The EU has long operated a Common Agricultural Policy
to support its farmers and has used several trade barriers to minimise the competition
faced by their domestic suppliers.
12.6.4
Financial factors
General price level of food
In general terms the price of food has been rising steadily over the last 20 years.
Currency impacts
As with many other commodities, the agricultural suite of products will be predominantly traded in US dollars. A substantial move in the exchange rate could have an
impact on both producers and consumers, depending on the direction of the move and
the location of the market participant.
140
120
100
80
60
40
20
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Consumer price indices; 2010 = 100; FAO
FIGURE 12.16 World food prices. Consumer price indices.
Source: FAO
418
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Investor activity
With the increasing use of commodity derivatives, large flows of speculative money
have moved into the futures market. For example, with the general increase in investor
interest in commodity indices such as the S&P GSCI, the ‘rolling’ of the index’s constituent futures (see Chapter 14) will influence the shape of the forward curve. This has
led some commentators to suggest that ‘speculation’ has caused the price of food for the
end consumer to be adversely impacted. This is a very contentious area, but there are a
couple of questions worth asking:
▪ How is speculation defined?
▪ Do speculators exploit trends or define them?
▪ Would the absence of speculators in the market decrease liquidity, therefore impacting the ability of commercial participants to hedge their exposures?
▪ What is the size of the futures markets relative to the underlying physical market?
▪ Since futures are typically semi-processed rather than the final products that are
consumed, what part of the physical supply chain is being impacted?
▪ Are the final prices paid by an end consumer impacted by futures activity? For
example, although Starbucks price their physical purchases using the coffee C contract, the cost of beans only makes up a small proportion of the final price paid by
the consumer.
▪ Although futures may have a short-term effect on prices, will they influence prices
over the longer term?
While researching this text the author came across examples of volatile markets
where there is either no or a very small futures market in operation. Examples include:
▪ Peanuts – US prices tripled in 2011 because of drought.
▪ Rubber – Prices increased by 75% in 2010. The rise was attributed to a fall in production in the previous year, rising demand, higher oil prices, and some investor
activity. At the time there was only a relatively small futures market for the product
based in Tokyo. Being able to attribute price movements to each of the underlying
factors would be nigh on impossible (Financial Times, 2010e).
▪ Barley – Again in 2010 the price of barley doubled in six weeks due to drought
conditions in Russia.
▪ Rice – In 2008, the price of rice increased from just over USD 300/tonne to over
USD 1,000/tonne over a couple of months. This led to extensive civil unrest as
approximately 3 billion people regard the commodity as a food staple (Financial
Times, 2008).
Commitment of traders report
The Commodity Futures Trading Commission (CFTC) produces a ‘Commitment of
Traders’ (COT) report on a weekly basis. According to the CFTC, ‘The COT reports
provide a breakdown of each Tuesday’s open interest for markets in which 20 or more
traders hold positions equal to or above the reporting levels established by the CFTC’.
Open interest is the total of futures and/or option contracts entered and not yet offset
419
Agriculture
by a transaction, exercise, or delivery. The total of all long open interest must equal the
total of all short open interest. Traditionally, the reports classified a trader as either being
‘commercial’ or ‘non-commercial’. Broadly speaking a commercial user will have some
form of underlying economic exposure that they are seeking to hedge using futures or
options. A non-commercial user would be more likely to use the contracts as a way of
profiting from an expected market movement. One aspect of this approach is that it
does not reveal the trader’s underlying motivation. So, a commercial user could execute
a transaction for speculative purposes rather than traditional hedging motivations.
To illustrate this, consider the following abbreviated report (Table 12.1) based on
the Soft Red Winter Wheat contract. Total open interest was 359,729 contracts.
From 2009, the CFTC supplemented this legacy method by introducing four categories of traders:
▪ Producer/merchant/processor/user – this designation was designed to capture those
participants engaged along the physical supply chain and would be more likely to
use futures for hedging purposes.
▪ Swap dealers – the CFTC define this category as ‘an entity that deals primarily
in swaps for a commodity and uses the futures markets to manage or hedge the
risk associated with those swap transactions. The swap dealer’s counterparties
may be speculative traders like hedge funds or traditional commercial clients that
managing risk arising from their dealings in the physical commodity’.
▪ Managed money – this could include commodity trading advisors (CTAs: individuals or organisations who provide advice and services relating to commodity
derivatives) and commodity pool operators (CPOs: individuals or organisations
who accepts money for the purpose of trading commodity derivatives).
▪ Other reportables – every other trader subject to the reporting requirements and not
captured by the other three categories is placed in this category.
Table 12.1 uses the same futures data as Table 12.2 but uses the amended reporting
format.
TABLE 12.1 Commitment of traders report – Soft Red Wheat futures traded at the Chicago
Board of Trade, 24 March 2020.
Entity
Position
Non-commercial
Long Short Spreads
Commercial
Long Short
Total
Long Short
Non-reportable
positions
Long
Short
Commitments 97,437 62,621 121,381 113,970 133,809 332,788 317,811 26,941
Percentage
27.1
17.4
33.7
31.7
37.2
92.5
88.3
7.5
of open
interest
Number of
97
75
108
77
108
239
244
traders
41,918
11.7
Note: A spread contract could be where a trader has both a long and short position in the same contract but
for different maturities.
Source: CFTC
TABLE 12.2 Disaggregated commitment of traders report – Soft Red Wheat futures traded at the Chicago Board of Trade, 24 March 2020.
Position
Commitments
Percentage of open
interest
Number of traders
Source: CFTC
420
Producer / merchant /
processor / user
Long
Short
Long
Swap dealers
Short Spreading
Managed money
Long Short Spreading
Other reportables
Long Short Spreading
28,192
7.8
104,769
29.1
69,832
19.4
13,094
3.6
15,946
4.4
62,852
17.5
42,709
11.9
88,589
24.6
34,585
9.6
19,912
5.5
32,792
9.1
51
87
20
10
17
59
33
58
38
42
50
Agriculture
421
These reports are often used as a measure of market sentiment and are often
reported with associated price movements. For example, the reports will show the
change in positions from the last reporting period and so significant swings could
indicate that the market is particularly bearish or bullish. The popular media will often
focus on activities of the more speculative categories of traders when trying to explain
price movements.
Emerging market wealth effects
No commodity price analysis would be complete without some mention of India and
China. As both countries continue to urbanise, it is likely that diets will evolve to include
more processed foods with higher sugar content and more protein. In addition, there is
likely to be an increase in demand for meat. The Economist (2007) points out: ‘calorie
for calories, you need more grain if you eat it transformed into meat than if you eat it
as bread: it takes three kilograms of cereals to produce a kilo of pork, eight for a kilo
of beef.’ So, the increase in wealth in these nations should lead to an increase in the
demand for corn as a feed for livestock.
A more general point about per capita incomes could be illustrated in the coffee
market. An increase in incomes may tempt drinkers away from instant coffee (perhaps
made with the cheaper robusta bean) towards more expensive drinks that will be
brewed using arabica beans.
12.7
EXCHANGE TRADED AGRICULTURAL AND ETHANOL
DERIVATIVES
Contract specifications
There are several competing exchange traded agricultural futures contracts, and
a sample of three popular commodities traded on the CME is listed in Table 12.3.
The exchange publishes a substantial amount of literature covering many aspects of
their use, which is available from their website (https://www.cmegroup.com/education
.html).
Popular trading strategies
It would be somewhat repetitive to reproduce all the futures content covered in other
chapters, so the intent in this short section is to highlight some popular trading strategies. Over the years several different agricultural futures strategies have evolved to replicate various underlying market relationships:
▪ ‘Hog crush’ trade – simulates the feeding process for hogs and consists of long positions in soymeal and corn against a short position in lean hogs.
▪ BBQ spread – replicating a possible barbecue menu would be coal or natural gas
futures against live cattle.
▪ Popcorn spread – corn against sugar.
422
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
TABLE 12.3 Futures contract specification for corn, soft red winter wheat, and soybeans. One
bushel of corn weighs approximately 56 pounds/25kg, while for soybeans and wheat a bushel is
taken to be 60 pounds/27kg.
Corn
Contract size
Deliverable
grades
Tick size
Price quote
Contract months
Last trading day
5,000 bushels
No. 2 yellow at par,
no. 1 yellow at 1 1/2
cents per bushel
over contract price,
no. 3 yellow at a
discount between 2
and 4 cents/bushel
depending on
broken corn,
foreign material,
and damage grade
factors.
1/4 cent/bushel (USD
12.50/contract)
Cents/bushel
Dec, Mar, May, Jul,
Sep
The business day
prior to the 15th
calendar day of the
contract month
Chicago Soft Red
Winter Wheat
Soybeans
5,000 bushels
Principally, no. 2 soft
red winter at par;
no. 1 soft red winter
at 3 cents/bushel
over contract price.
5,000 bushels
No. 2 yellow at par,
no. 1 yellow at a
6 cent/bushel
premium, no. 3
yellow at a 6 cent/
bushel discount
1/4 cent/bushel (USD
12.50/contract)
Cents/bushel
Dec, Mar, May, Jul,
Sep
The business day
prior to the 15th
calendar day of the
contract month
1/4 cent/bushel (USD
12.50/contract)
Cents/bushel
Jan, Mar, May, July,
Aug, Sep, Nov
The business day
prior to the 15th
calendar day of the
contract month
Source: CME Group
The CME trades the entire soy complex and from this, a range of strategies has
evolved. The ‘crush spread’ is the expected gross margin from the processing of soybeans
and is a dollar amount determined by the price of soybeans relative to the combined
price of soybean meal and soybean oil. The trading of the spread is predicated largely on
expected price movements of the soybeans against the meal and the oil components of
the crush. The trade is constructed using futures by the purchase (sale) of the soybean
contract and the sale (purchase) of both the soybean oil and soybean meal contracts.
Christian (2006) argues that one of the motivations for trading this price spread is since
the oil content and the yield will vary between crops and harvests. A typical bushel
of soybeans weighs around 60 pounds (just over 27 kg), which when crushed might
yield 11 pounds (5 kg) of soybean and around 44 pounds (20 kg) of soybean meal. About
5 pounds (about 2.26 kg) will go to waste.
If meal demand is high and oil demand is not, then processors may decide to switch
the balance of production between the two products. One popular way of trading this
relationship is by means of options on the spread between the prices. The fundamentals
of spread options were covered in Chapter 1.
Agriculture
423
‘Chocfinger’ – a case study on short squeezes
In June 2010, press reports surfaced that the July cocoa futures price had risen above
GBP 2,500 for the first time in 33 years. This was initially attributed to a poor harvest
in the Ivory Coast earlier in the year, which had led to a fall in supply. The cocoa plants
were characterised as being old and prone to disease and with demand having outstripped supply for five years, prices had been rising steadily. As events started to unfold
it became clear that the ‘open interest’ in the July contract was significantly higher than
normal, raising suspicions that prices were being driven higher by speculative activity,
with perhaps one or two dominant players having large, long positions.
Initially, the impact on market participants was twofold. As prices increased, cocoa
bean consumers started buying greater volumes of July expiry call options. But this had
an unintended impact on market prices because of the hedging activities of the banks
that had sold the options. The sale of call options by the banks would expose them
to potential open-ended losses if prices continued to rise. The classic hedging strategy
to mitigate this is for the banks to ‘delta-hedge’ their directional exposure by buying
futures. This offsetting position should offer some protection against rising prices. However, this had something of a feedback effect in that the purchase of these futures would
push prices even higher.
The second impact was the hedging activity of cocoa processors who transform the
beans into cocoa butter and liquor, which are intermediate products in the production of
chocolate. The processors are buyers of cocoa beans and sellers of the intermediate butter and liquor products. These processors would typically sell cocoa futures to hedge the
sale of their intermediate products against a possible fall in price. This strategy assumes
that the price of the cocoa future moves broadly in line with the prices of the intermediate products. Typically, the processors will have an ongoing hedging requirement
and would not be either willing or able to take the contract to final delivery. Since the
contract was physically settled, the processors would be required to either deliver what
inventory of beans they held or procure additional physical supplies from the market.
Since the end products they were trying to hedge (i.e. butter and liquor) were not eligible for delivery into the futures contract, these participants would typically roll their
exposure. The processors would sell (say) the July contract and shortly before its expiry,
close out the exposure by taking an offsetting long position. They would then immediately sell a longer-dated maturity such as September to maintain their exposure to
the market. This short September position would again be held until close to maturity
where the roll strategy would be repeated; buy to close out the September contract, sell
December to re-establish their directional hedge.
As prices continued to increase, the market turned into backwardation with the
July contract trading GBP 156.00 higher than the September maturity. This backwardation was signaling a short-term excess of demand over supply. This prompted some
market participants to complain to the exchange (at the time NYSE LIFFE, now the ICE)
that the market was being manipulated. One market participant commented: ‘the shorts
are the ones complaining . . . if they believe it is all just speculation, they can call the
bluff and deliver at expiry physical cocoa and scare the speculators’, (Financial Times,
2010g). As demand continued to increase, the price spread between the first and second
‘prompt’ futures increased to GBP 300.00/tonne as the contract approached expiry.
424
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
The July 2010 contract finally matured on the 15 July with the price hitting GBP
2,750/tonne. It then transpired that a UK hedge fund, Armajaro, had taken physical
delivery of 240,100 tonnes of cocoa. As a result, Anthony Ward, the owner of the fund
was dubbed ‘Chocfinger’ by the press. Since one future was 10 metric tonnes, the hedge
fund had gradually built up a long position of 24,010 futures, with an at expiry market
value of about GBP 660 million. Press reports suggested that the long position in the July
2010 contract had been initiated in October 2009 and had been added to on an incremental basis. This was the largest physical delivery seen on the exchange for 14 years
and represented about 7% of annual global production. Indeed, this was not the first
time that Ward had taken large positions in cocoa futures. In 1996 he used the futures
market to take delivery of 300,000 tonnes of beans (about 10% of global production at
the time) and in 2002 a similar strategy saw him take delivery of 148,000 tonnes.
Although the exact motivation of the hedge fund has never been revealed, their
strategy appeared to suggest a belief that prices would continue to increase and that
the October 2010 crop would be poor, following the trend of previous years. By early
September, reports emerged that the October crop in the Ivory Coast would be better
than expected and that the market would return to a surplus of supply for the first
time in several years. As a result, futures prices for December delivery started to fall
below GBP 2,000. It is unclear from reports in the public domain the extent to which
Armajaro’s strategy was profitable. Some commentators suggested that the hedge fund
had in place several offtake agreements with chocolate producers so it seems likely
that these deliveries would have taken place sometime in the third quarter of the year.
Most chocolate producers tend to increase production from about September to meet
Christmas demand. By November 2010, the price of cocoa beans had fallen to GBP 1,770
and since some articles suggested that Armajaro had paid on average between GBP
2,100 and GBP 2,200/tonne, the fund may have been facing a loss on any remaining
inventory. Additionally, the cost of financing a long position held in a warehouse would
be significant at about USD 7–10 million a month (e.g. interest costs, warehousing rent).
Armajaro started to sell its remaining physical position in September 2010 and completed it by December of the same year. Interestingly, as the December 2010 contract
expired, nearly 110,000 tonnes of cocoa were physically delivered, the largest amount
for the year-end contract for 10 years. It was suggested that Armajaro made most of
the deliveries. It is perhaps plausible to suggest that after taking physical delivery in
mid-July, the hedge fund may have established a short December futures position at
what was then a higher price than the final settlement price. Focus then turned to the
buyers of this December delivery, which turned out to be principally two large traders.
Arguably, they were concerned with rising prices as their purchases coincided with
political unrest in the Ivory Coast, which led to a virtual close of the cocoa industry.
The incident does highlight several interesting points:
Difficulty in defining market manipulation – Although the regulatory definition of
market manipulation is couched in somewhat formal legal language some, but not all,
of the main principles involved are:
▪ Providing misleading signals that may impact the supply or demand of a product.
▪ Buying and selling at off market prices.
▪ Entering a transaction that employs some form of deception.
Agriculture
425
▪ Knowingly disseminating false or misleading information to impact either supply
and demand or price.
▪ Inputting false or misleading information into the process of compiling a benchmark index.
▪ Trading at opening and closing prices, which may mislead investors.
▪ Voicing an opinion in the media in which may benefit an existing position without
declaring the conflict of interest.
One definition that is worth reproducing in full that is relevant to the case is (European Union, 2014):
‘The conduct by a person, or persons acting in collaboration, to secure a dominant position over the supply of or demand for a financial instrument, related
spot commodity contracts . . . which has, or is likely to have, the effect of fixing,
directly or indirectly, purchase or sale prices or creates, or is likely to create,
other unfair trading conditions’.
The allegation of market manipulation made by the market participants in relation
to the July 2010 contract was investigated by the exchange that argued that the trading
activity was not designed to specifically to distort the price of the contract. The exchange
pointed out that since the hedge fund had taken delivery of the physical product they
had made use of one of the primary functions of the exchange in that it is willing to act
as the buyer and seller of last resort for physical supplies. Admittedly, the EU definition
of market manipulation supersedes the 2010 incident, but it illustrates not only the difficultly of applying regulations that are based on principles rather than strict rules, but
also how regulation will often evolve to address developments in the market.
Liquidity – Typically when buying and selling most financial products on an
exchange there may be an expectation that a position can be easily reversed. Consider
the situation of the processors who were structurally short futures and had become
used to the idea of buying back the position to avoid physical delivery. No doubt as
the contract’s maturity approached, they started to realise it was nearly impossible to
find a willing seller as the dominant participant was intent on taking physical delivery.
Presumably, to buy back their position so as to avoid physical delivery, they were
willing to bid prices higher and higher, further contributing to the price increase.
Proxy hedging – Although information in public domain is somewhat sketchy,
there is a possibility that the rise in price and the associated movement of the curve
into backwardation could impact the processors in two different ways. If they had
not hedged their purchase of beans, the rise in futures prices may well have led to
an increase in their physical purchase costs. Secondly, the losses incurred from their
short futures position could reduce their overall profitability as the income from
the sale of the butter and liquor products would not automatically increase. Like
many market participants, they were caught between two different markets and therefore had an exposure to the price spread. There was no exchange traded solution that
allowed them to manage their exposure to a fall in the price of their liquor and butter
products and so they used the cocoa bean futures contract as a proxy. Arguably, an
over-the-counter structure would have provided a better solution although it would
426
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
leave their counterparty with the ‘proxy risk’. This risk would have to be priced into the
contract making it less attractive but arguably resulting in a more effective hedge. This
is a similar problem to the jet fuel hedging example illustrated in Chapter 6.
12.8
OVER-THE-COUNTER AGRICULTURAL DERIVATIVES
Anecdotally, the maturity of hedges instigated by participants in the food sector will
often be shorter than in other industries. As production can vary from year to year due to
factors such as weather or disease, companies are very often reluctant to hedge beyond
a 12–18-month period (Financial Times, 2010d).
Swaps
OTC contracts such as swaps exist for most agricultural products and often banks
can offer swaps on products where no liquid future exists. This idea was discussed in
Chapter 6 with respect to jet fuel. Like examples in previous chapters, swaps could be
used for several reasons that include:
▪ Transforming an underlying price exposure from floating to fixed or vice versa.
▪ Taking a directional view on the price of the commodity without having to take
physical delivery.
The swap could either cover a single period in the future or involve multiple periodic cash flows. A hypothetical term sheet may look as follows:
Commodity:
Trade date:
Effective date:
Termination date:
Notional amount:
Fixed price payer:
Fixed price:
Floating price payer:
Reference price:
Corn
August
Three months from trade date (i.e. 1 November)
1 month from effective date (i.e. 30 November)
700,000 bushels
Counterparty
USD 3.5000/bushel
Bank
Corn – CME; the arithmetic average of the daily settlement
price of the first nearby month futures contract over the
period of the swap
OTC swap contracts for wheat will be very similar in structure and typically settle
against the CME futures price. Similar to the base metals market, commercial contracts
may well be linked to a futures price as this allows the hedger to ensure that the price on
which the physical contact is based is the same as that used for the applicable hedging
instrument.
427
Agriculture
Fixed price spread
Client
Bank
Monthly average
of spread
Soybean oil price
Diesel
FIGURE 12.17 Vanilla swap referencing the spread between soybean oil and diesel.
Vanilla swaps – exotic requirements
Biodiesel is a fuel made from a variety of different sources, which include vegetable oils
such as palm oil. It is more environmentally friendly as in its pure form it is non-toxic
and biodegradable. Like ethanol, biodiesel can be blended in certain proportions with
conventional diesel fuel without any need to modify the engine. Consequently, biofuel producers are faced with a spread exposure as part of the business. They consume
agricultural products and have their outputs linked to an oil product. Although this
spread exposure may appear to be ‘exotic’ it would be possible to structure a vanilla swap
that transforms a variable spread exposure to a fixed value. Figure 12.17 illustrates the
concept at a high level.
In this example the client is a biofuel producer who has an expense related to soybean oil but receives income from the sale of their biodiesel fuel. In the swap transaction
they pay away this ‘floating’ spread. This spread could be calculated as the average of
daily prices observed over a given period such as a month. In return they would receive
a fixed cash flow. So this would be beneficial if the price of soybean oil increases for a
prolonged period relative to the cost of diesel. There is something of a basis risk here
as the producer is selling biodiesel but exchanging cash flows whose values may well
reference a liquid futures price such as ultra-low sulfur diesel.
Options
Vanilla options on agricultural products are well established and readers interested in
different structures are referred to the examples documented elsewhere in the book (e.g.
Chapter 4 for gold and Chapter 5 for base metals).
Options on spreads
Biodiesel producers will often monitor certain spreads that measure the margins
between biodiesel and diesel fuel. These spreads include:
▪ POGO (Palm oil minus gas oil).
▪ BOHO (Soybean oil minus heating oil).
▪ BOGO (Soybean oil minus gas oil).
428
COMMODITY DERIVATIVES: MARKETS AND APPLICATIONS
Suppose that a biodiesel producer sees that palm oil prices are declining, making
the commodity attractive for blending into diesel. They see that the current POGO
spread is trading at −USD 50.00/MT (i.e palm oil prices are lower than gas oil), which
is below the long-term average they have calculated as being + USD 100/MT. However,
they realise that this may not persist as other producers would also likely take advantage
of the situation